Skip to main content

Full text of "Mangrove Forest Management Guidelines Fao Forestry Paper 117"

See other formats

Mangrove forest 
management guidelines 

Forest Resources Development Branch 
Forest Resources Division 
FAO Forestry Department 

The designations employed and the presentation of material in this 
publication do not imply the expression ot any opinion whatsoever on 
the part of the Food and Agriculture Organization of the United 
Nations concerning the legal status of any country, territory, city or 
area or of its authorities, or concerning the delimitation of its frontiers 
or boundaries. 

ISBN 92-5-103445-1 

All rights reserved. No part of this publication may be reproduced, stored in a 
retrieval system, or transmitted in any form or by any means, electronic, mechani- 
cal, photocopying or otherwise, without the prior permission of the copyright owner. 
Applications for such permission, with a statement of the purpose and extent of the 
reproduction, should be addressed to the Director, Publications Division, Food and 
Agriculture Organization of the United Nations, Viale delle Terme di Caracalla, 
00100 Rome, Italy. 

FAO 1994 


Mangroves are unique ecosystems. As a source of renewable resources, 
they are second to none in terms of its natural productivity and the wide 
range of goods and services they provide on a continuing basis. The economic 
potential of mangroves stems from three main sources, namely, forest products, 
estuarine and near-shore fisheries, and ecotourism. In addition, mangroves 
play a pivotal role in coastal protection and maintenance of habitats for a 
large range of common, threatened and endangered species. 

Due to pressures from growing populations, which lead to changes in land 
use and over-utilization of the resources, mangroves are being rapidly 
depleted and degraded. A balance needs to be struck between meeting 
increasing present-day needs, on the one hand, and conserving the 
environmental support system provided by the mangrove forests, on the other. 

Growing awareness of the protective, productive and socio-economic 
functions of tropical mangrove ecosystems, and of the consequences of their 
deterioration, has highlighted the need for the conservation and sustainable, 
integrated management of these valuable resources. Given their multiple-use 
potential, it is imperative that the management of mangrove based terrestrial 
and aquatic ecosystems be undertaken within the context of integrated coastal 
area management planning. 

In many countries, much of the basic information needed for development 
and execution of management plans in mangrove forests, is presently not 
available. These guidelines present a synthesis of mangrove forest management 
systems which have been developed and successfully used in a number of 
countries and regions, and examine experiences gained and lessons learned. 

The guidelines focus specifically on the management of the forest 
resources contained in the mangrove ecosystem, including wood and non-wood 
forest products. A chapter on mangrove ecology is included in order to ensure 
adequate understanding of the dynamics of these ecosystems, as a basis for 
their conservation and sustainable use. 

The guidelines further include chapters on inventory and assessment of 
mangrove resources, and on traditional and potential uses of products provided 
by them. Environmental impact assessment is finally reviewed, and conclusions 
and recommendations are given to summarize the findings in the document. 

It is hoped that the present guidelines will contribute towards improved 
understanding of the mangrove ecosystem and the natural renewable resources 
contained in them, and that they will aid in the development and 
implementation of integrated, multiple-use management plans to ensure the 
sustainability of these resources, now and in the future. 

J . P . Lanly 

Director f 
Forest Resources Division 
Forestry Department 


In this study of mangrove forest management, the section on Assessment 
of Mangrove Resources is based on the Manual on Mapping and Inventory of 
Mangroves prepared by D. Benessalah and issued as a draft miscellaneous paper 
by the FAO Forestry Department in 1988 (FO:MISC/88/l) . Most of the 
information contained in the sections on The Ecological Basis for Mangrove 
Management Planning; Traditional and Potential Uses of Mangrove Resources; and 
Sustainable Resource Management was provided by P.W Chong in a study prepared 
for the FAO in 1993. A case study on Multiple-use Management of the 
Sundarbans Forest in Bangladesh was prepared by M.Z. Hussain of IUCN. J. 
Troensegaard contributed the chapter on Environmental Impacts of Mangrove 
Management and provided valuable comments on the first draft of this paper. 
M. Loyche Wilkie consolidated the above studies and comments to produce the 
present synthesis. Final editing was by P. Vantomme. FAO is indebted to the 
above authors and to the numerous individuals and institutes, which provided 
the basic information contained in this document. 

FAO sincerely thanks the 'Tropical Forestry Program' of the U.S. Forest 
Service, United States Department of Agriculture, for their support in the 
publication of the present document. 


GLOSSARY : . . . 







2.1 Definition of mangroves . 5 

2.1.1 Floristics 5 

2.1.2 Mangrove taxonomy 6 

2.2 Biogeography 6 

2.3 Mangrove ecology 11 

2.4 Climate 12 

2.4.1 Temperature 13 

2.4.2 Winds and storms 13 

2.4.3 Rainfall 14 

2.4.4 Life Zones (Zonas de Vida) 14 

2.5 Edaphic factors 14 

2.5.1 Mangrove geomorphology 14 

2.5.2 Salinity 21 

2.5.3 Other edaphic factors 23 

2.6 Mangrove vegetation 25 

2.6.1 Vegetal formations and communities ..... 25 

2.6.2 Zonation and inundation . 27 

2.6.3 Succession - ecological aspects 31 

2.7 Mangrove fauna 34 

2.7.1 Wildlife 34 

2.7.2 Avifauna 37 

2.7.3 Aquatic resources 37 

2.8 Trophic relationships in mangrove ecosystems ... 39 

2.8.1 The food web 39 

2.8.2 Primary wood production 41 

2.8.3 Secondary production 42 

2.8.4 Keystone species 43 

2.8.5 Management implications 44 




3.1 Mangrove -based products and services 45 

3.2 Utilization of wood products 47 

3.2.1 Timber 48 

3.2.2 Charcoal 49 

3.2.3 Firewood 53 

3.2.4 Fishing stakes/poles 55 

3.2.5 Pulp 55 

3.2.6 Tannin ; 55 

3.3 Utilization of non-wood resources 56 

3.3.1 Nipa palm 57 

3.3.2 Apiculture 59 

3.3.3 Wildlife 61 

3.3.4 Capture fishery 63 

3.3.5 Mariculture 66 

3.3.6 Salt production 72 

3.3.7 Agriculture on mangrove soils 74 

3.4 Services provided by mangroves 78 

3.4.1 Coastal protection 78 

3.4.2 Recreation and ecotourism 80 

3.5 Integration of uses 82 

3.5.1 Integrated coastal area planning 82 

3.5.2 Land use issues and conservation 83 

3.6 The socio-economic value of mangroves- 86 



4.1 Planning levels 89 

4.1.1 Land-use planning 89 

4.1.2 Forest planning 91 

4.1.3 Monitoring & Evaluation 92 

4.2 How to obtain the information needed 92 


5.1 Choice of sensor 93 

5.1.1 Application of aerial photography to mangrove 
areas 95 

5.1.2 Application of satellite imagery to mangrove 
areas 98 

5.1.3 Application of radar imagery to mangrove 
areas 103 

5.1.4 Comparison between major sensors 107 


6.1 Classification of mangroves Ill 

6.2 Survey designs 113 

6.2.1 National level surveys 114 

6.2.2 Forest management surveys 117 

6.2.3 Operational level surveys 122 

6.3 Mapping of mangroves 122 

6.4 Area estimation 123 


7.1 Forest resource assessment in mangrove areas . . , 128 

7.1.1 Volume estimation from remote sensing imagery 128 

7.1.2 Volume estimation with limited field sampling 130 

7.2 Forest inventories of mangroves 130 

7.1.1 Sampling designs 130 

7.1.2 Sampling intensity 134 

7.1.3 Sampling unit shape and size 135 

7.1.4 Continuous forest inventory (CFI) 136 

7.3 Forest mensuration in mangroves 137 

7.3.1 Tree characteristic measurements 137 

7.3.2 Volume determination 138 

7.3.3 Growth determination 139 

7.4 Presentation of results 139 



8.1 Planning for multiple use management 142 

8.2 Supply and Demand 143 

8.3 People's Participation 143 

8.4 Policy framework 144 

8.5 Principles of Planning 144 


9.1 Basic Planning Steps 148 

9.2 Defining the management area and the duration of the 

plan 150 

9.3 Collection of basic information 151 

9.3.1 Data types 151 

9.3.2 Resource data 151 

9.3.3 Operational data 152 

9.3.4 Utilisation data 152 

9.3.5 Socioeconomic data 153 

9.3.6 Institutional data / 153 

9.4 Management goals and objectives . . 154 

9.5 Plan strategy 156 

9.6 The concept of sustainable ecosystem development . 157 

9.7 Division of area 158 

9.8 Preparation of the working plan 159 


10.1 Choice of silvicultural system 160 

10.1.1 Clear-felling systems 160 

10.1.2 Selection systems 162 

10.1.3 Shelterwood systems 164 

10.2 Choice of species 164 

10.3 Natural regeneration 164 

10.3.1 Seed sources for natural regeneration .... 165 

10.3.2 Retention of standards (seed-bearers) . . . 165 

10.4 Regeneration stocking 166 

10.4.1 Inadequate regeneration 167 

10.4.2 Regeneration classes 167 

10.4.3 Regeneration stocking adequacy standards . . 167 

10.4.4 Linear regeneration sampling 168 

10.4.5 Effective stocking 169 

10.5 Artificial regeneration 169 

10.5.1 Phenology . . 169 

10.5.2 Collection of propagules 171 

10.5.3 Site preparation 171 

10.5.4 Nursery operations 171 

10.5.3 Planting 171 

10.5.6 Afforestation of newly formed mudflats . . . 173 

10.5.7 Reforestation of degraded areas 175 

10.5.8 Refilling 176 

10.6 Weed control 176 

10.7 Disease and pest control 176 

10.8 Thinning 178 

10.9 Choice of rotation 178 

10.9.1 Selection of rotation 180 

10.10 Conservation and protection areas 180 

10.10.1 Genetic biodiversity 180 

10.10.2 Erosion control 181 

10.10.3 Avifauna 181 

10.10.4 Other wildlife 181 

10.10.5 Fisheries 181 

10.10.6 Recreation and education 181 


11.1 Estimation of yield 183 

11.1.1 Rates of growth 183 

11.1.2 Yield and production 184 

11.1.3 Effective logging area 185 

11.1.4 Estimation of product mix 186 

11.2 Forest yield regualation 187 

11.2.1 Determination of the annual cut 187 

11.2.2 Regulation in even-aged stands 189 

11.2.3 Regulation of uneven-aged stands 189 

11.2.4 Control of removals 190 


12.1 The felling plan 192 

12.1.1 Felling strips 192 

12.2 Choice of harvesting system 193 

12.2.1 Wheelbarrow method 194 

12.2.2 Tramway 194 

12.2.3 Canals 194 

12.2.4 High- lead cable system 197 

12.2.5 Portable cable winch 198 

12.2.6 Manual extraction 198 

12.3 Logging damage 199 


13.1 The operational plan 200 

13.2 Plan monitoring 200 

13.2.1 Records to be kept 200 

13.2.2 Supervision and control 200 

13.2.3 Costs and revenues 201 

13.3 Plan evaluation and revision 202 


14.1 Environmental concerns 203 

14.2 Management activities and their environmental 
impacts 203 


15.1 Forest policy and legislation 206 

15.1.1 Land use policy 206 

15.1.2 Multiple-use .concept 206 

15.1.3 Legislation 207 

15.2 Resource inventory 207 

15.3 Establishment of permanent mangrove reserves . . . 207 

15.3.1 Classification of forest use categories . . 208 

15.3.2 Land tenure and usufruct 208 

15.4 Forest Service 208 

15.4.1 International technical assistance 208 

15.5 Management, silviculture and utilization 209 

15.6 Socioeconomic and financial aspects 210 

15.7 Social forestry, extension and demonstration . . . 210 

15.8 Applied research 210 

15.8.1 Environmental Impact research 211 

15.8.2 Socioeconomic studies 211 

15.8.3 Demand for mangrove products 211 

15.8.4 Evaluation of mangrove management 
policies/programmes 211 

15.8.5 Ecological and silvicultural studies .... 211 

15.8.6 Zonation and site classification 212 

15.9 Conservation, wildlife and tourism 212 

















Table 12.2 Comparative snigging costs using manual and 

winch methods 198 


Fig. 2.1 Rhizophora apiculata with ripe propagules 8 

Fig. 2.2 Nypa fruticana with fruits 8 

Fig. 2.3 Avicennia africana with pneumatophores 9 

Fig. 2.4 Bruguiera gymnorrhiza with flowers 9 

Fig. 2.5 Generalized distribution of mangroves 10 
Fig. 2.6 Schematic representation of an idealized 

estuary - coastal lagoon 18 
Fig. 2.7 Distribution of tidal types in Southeast Asia 21 
Fig. 2.8 Mangrove community types 26 
Fig. 2.9 Vegetational changes and average value of water 
properties along a transect line at different 
distances from the waterway in Thailand ' 28 
Fig. 2.10 Proboscis monkey feeding on Rhizophora propagule 34 
Fig. 2.11 The spotted deer (Axis axis) in the Sundarbans- 36 
Fig. 2.12 The saltwater crocodile (Crocodilus porosus) 36 
Fig. 2.13 Large Sandplover 37 
Fig. 2.14 Generalized food web in mangrove ecosystem 40 
Fig. 2.15 Some ecological interactions between various land- 
uses and economic activities in mangrove areas 44 

Fig. 3.1 Masonry kiln for charcoal making in Indonesia 51 
Fig. 3.2 Charcoal making by the earth mound method in 

Guana 1, Cuba 51 
Fig. 3.3 Rhizophora firewood bundled with split mangrove 

stilt-roots 54 

Fig. 3.4 Bark collection in Southern Vietnam 54 
Fig. 3.5 Collection of Nypa leaves in the Sundarbans, 

Bangladesh 58 
Fig. 3.6 Shingles (Atap) made from Nypa leaves in 

Sumatra, Indonesia 58 

Fig. 3.7 Bark hive placed on tree in Africa 59 

Fig. 3.8 Marine Green turtle 62 

Fig. 3.9 Setting traditional fish traps in Matang, Malaysia 65 
Fig. 3.10 Boat load of Blood cockles (Anadara grranosa) , Matang 65 

Fig. 3.11 Cage culture, Matang, Malaysia 69 

Fig. 3.12 Oyster Culture along a mangrove creek, Sierra Leone 69 

Fig. 3.13 Shrimp 70 
Fig. 3.14 Clearing of mangrove area for shrimp ponds, Malaysia 71 

Fig. 3.15 Monitoring the shrimp production, Malaysia 71 
Fig. 3.16 Lands suitable for pond construction in relation to 

tidal elevations in the Philippines 73 
Fig, 3.17 Leaching of mangrove soil with salt water, 

Sierra Leone 75 
Fig. 3.18 Boiling of the brine solution in big pans, 

Sierra Leone 75 

Fig. 3.19 Effect of road on mangroves, Cuba 84 

Fig. 3.20 A shrimp-tree farming model 85 

Fig. 5.1 & Aerial photos showing mangrove vegetation types 

Fig. 5.2 in Mexico 96 

Fig. 5.3 IR colour aerial photos showing mangrove vegetation 

types 97 
Fig. 5.4 & 

Fig. 5.5 Examples of Landsat MSS imagery of mangrove areas 99 
Fig. 5.6 False colour composite of a SPOT image after band 

correlation 101 
Fig. 5.7 Colour composite of SPOT image and corresponding 

sketch map showing mangrove species 102 
Fig. 5.8 Colour composite and map showing mangrove plantations 102 

Fig. 5.9 Air photo showing cloud cover over a mangrove area 104 

Fig. 5.10 A radar image of the coastal zone of Colombia 105 


Table 2.1 

Table 2.2 

Table 2.3 

Table 2.4 

Table 2.5 

Table 2.6 

Table 2.7 

Table 2.8 
Table 2.9 

Table 3.1 

Table 3.2 

Table 3.3 

Table 3.4 

Table 3.5 

Table 3.6 

Table 3.7 

Table 3.8 

Table 4.1 

Table 5.1 
Table 5.2 
Table 6.1 
Table 6.2 
Table 6.3 

Table 7.1 
Table 7.2 
Table 8.1 

Table 10.1 
Table 10.2 

Table 11.1 
Table 11.2 
Table 11.3 

Table 11.4 
Table 11.5 

Table 11.6 
Table 11.7 

Number of mangrove species 6 

A list of mangrove species world wide 7 

Approximate mangrove areas in various countries 11 

Global warming scenarios 15 

Effect of salinity on maximum growth of mangroves 22 

Salinity-growth relationship for penaeid shrimp 23 

Inundation classes according to Watson's and 

De Harm's classification 30 

New and Old World species occurrence according to 31 

inundation classes 

Above ground biomass of Rhizophora apiculata 41 

Mangrove -based products 46 

Wood density for selected species 48 

Yield of charcoal for selected species 49 

Greenwood input and charcoal output per burn 50 

Fuel wood/ fish ratios in fish- smoking - Sierra Leone 53 

Heating value of selected mangrove species 53 

Capture fishery production in relation to mangroves 63 

Rice cultivation systems 77 

Examples of information needs at different planning 

levels 90 

Comparative photo -interpret at ion of panchromatic 

photography and SIR- A images 107 

Relative advantages of major sensors 108 

Relative importance of forest survey elements in 

mangrove forests 113 
The influence of tidal level on aerial photography 

coverage 118 

Type of information and map scale requirements for 

different application levels 122 

Percent of total baric volume for Rhizophora mangle/R. 

harrisonii trees of different diameter class 139 

Stand table for Rhizophoras/Pelliciera rhizophorae 141 

Forest management planning levels and 

responsibilities 147 

Average stand density and mortality rate for 

Rhizophora apiculata 168 
Minimum seedling regeneration stocking for different 

sampling quadrat size (RC I) 168 

Diameter growth rates of R. apiculata trees by 

diameter size classes (1920-81) 183 

Diameter increment of R. apiculata by crown classes 

in Pulau Kecil 183 

DBH (overbark) increments for Rhizophoras in the 

pure and overlap zones along the North Pacific Coast, 

Costa Rica 184 

Volume in m'/ha per diameter class, Playa Garza, 

Costa Rica 184 

Growth and Yield of Rhizophora apiculata plantations 

of different ages and sites in Yeesarn, 

Samut Songkram Province, Thailand 185 

General specifications for selected mangrove -based 

Products in Costa Rica 186 

Estimated products range based on diameter classes 

and merchantable height (10 cm overbark) 186 

Fig. 5.11 A radar image showing mangrove formations 106 
Pig. 5.12 A black and white Shuttle Imaging Radar (SIR-B) image 109 

Fig. 5.13 Enhanced colour composite of a SIR-B radar image 109 

Fig. 6.1 Stereogramme showing mangrove forest types 121 
Fig. 6.2 A small scale land-use map based on digital Landsat 

image classification with a smoothing process 124 
Fig. 6.3 A medium scale land-use and forest type map based 

on a SPOT image 125 
Fig. 6.4 A large scale mangrove management map showing 

compartments and logging areas 126 

Fig. 7.1 An illustration of a systematic strip lay-out 132 

Fig. 7.2 An illustration of a line plot lay-out 132 
Fig. 7.3 An illustration of a stand table for a 

'normal forest' 141 

Fig. 10.1 Dense natural regeneration of Rhizophora spp. 170 

Fig. 10.2 Artificial regeneration of Rhizophora apiculata 170 
Fig. 10.3 Rhizophora racemosa propagules ready to be planted 172 
Fig. 10.4 A mangrove nursery with R. apiculata and 

B. gymnorrhiza 172 

Fig. 10.5 Planting of mangroves by school children 174 

Fig. 10.6 Rhizophora racemosa 2 years after planting 174 

Fig. 10.7 Acrostichum infested area, Matang, Malaysia 177 

Fig. 10.8 Defoliation of pristine R. racemosa stand 177 

Fig. 10.9 Rhizophora apiculata stand after first thinning 179 
Fig. 10.10 Matura Rhizophora apiculata stand ready for 

final felling 179 

Fig. 10.11 The 'Virgin Jungle Reserve 1 , Matang, Malaysia 182 

Fig. 10.12 Walkways on stilts 182 

Fig. 11.1 Life-pattern of an even-aged stand 189 

Fig. 11.2 Life-pattern of an uneven-aged stand 190 

Fig. 12.1 Extraction of wood by wheelbarrow 195 

Fig. 12.2 Loading of billets unto the boat 195 

Fig. 12.3 Man-made canal in Guanal, Cuba 196 


5.1 Resolution requirements and survey levels 94 

6.1 A schematic classification of mangrove areas 112 

6.2 Sequence of a cartographic survey based on (a) aerial 

photos and (b) Satellite images 115 


Box 2.1 The Delta building process 16 

Box 2.2 The acid sulphate problem 24 

Box 2.3 Caimans as keystone species 43 

Box 3.1 Notes on good bee colony management 61 
Box 3.2 An economic analysis of mangrove reclamation for 

agriculture in Fiji 76 

Box 3.3 Coastal protection in Guyana 79 
Box 3.4 Protection of coastal dykes by mangroves in Vietnam 81 

Box 3.5 Guidelines on planning ecotourism in mangroves B2 

Box 3.6 Integration of uses in mangroves in Vietnam 85 

Box 3.7 Socio-economic value of mangroves in Fiji 87 

Box 7.1 Determination of bark volume 140 

Box 9.1 Management objectives 154 

Box 9.2 Dat Mui Forest Enterprise - management goals and 

objectives 155 

Box 9.3 Plan strategy for a forest enterprise in Vietnam 156 

Box 9.4 Plan strategy for the Sierpe- Terraba mangroves in 

Costa Rica . . 157 

Box 9.5 Definition of a 'normal forest' 157 

Box 10.1 Silvicultural systems 161 

Box 10.2 Advantages and disadvantages of Clear-felling Systems 162 

Box 10.3 Advantages and disadvantages of Selection Systems 163 

Box 10.4 Advantages and disadvantages of Shelterwood Systems 163 

Box 10.5 Advantages and disadvantages of Natural Regeneration 165 

Box 10.6 Criteria for the selection of standards (seed-bearers) 166 

Box 10.7 Regeneration classes 167 

Box 10.8 Aerial seeding of mangroves 173 

Box 11.1 Estimation of product-mix in Costa Rica 186 

Box 12.1 Guidelines for canalization with explosives 197 


Afforestation The planting of trees in unfbrcstcd areas. 

AllochthoHOUs Of inhabitants from the outside (as opposed to Autochtonous). 

Alluvium Material carried in suspension by rivers and deposited in sluggish water beyond the influence 

of the swiftest current. Some of die worlds most fertile soils are alluvial. 

Aquifer A layer of permeable rock, sand or gravel that absorbs water and allows it free passage 

through the intervening spaces of the rock. When the underlying rock is impermeable an 
aquifer acts as a ground water reservoir. 

Autochtonous Of original inhabitants. 

Backplain That part of a river floodplain, between the levee and the backswamp, normally flooded for 
several months each year. 

Backswamp The lowest part of a river floodplain experiencing prolonged flooding. 
Benthos Community of organisms inhabiting the bed of a water body. 

Biogeography The study of the distribution of animals and plants. 

Bole Merchantable part of the stem from the stump cross-section to the merchantable limit which 

is defined as the crown point or a certain upper diameter. 

Buffer zone A zone, peripheral to a national park or equivalent reserve, where restrictions are placed upon 
resource use or special development measures are undertaken to enhance the conservation 
value of the area. (ICVN,1991) 

Bund A dike or embankment. 

Cover Type (more commonly Forest Cover Type) A descriptive term used to group stands of similar 

characteristics and species composition (due to given ecological factors) by which they may 
be differentiated from other groups of stands. 

Conversion A change from one silvicultural system or species to another. 

Conversion forest Forests assigned for conversion to agriculture (Agri-conversion) or other non-forestry use. 

Deforestation The clearing of forests and the conversion of land to non-forest uses. 

Degradation (more commonly Forest Degradation) Biological, chemical, an physical processes that result in 
loss of productive potential of natural resources in areas that remain classified as forests. 
Degradation may be permanent, although some forests may recover naturally or with human 

Depletion Reduction in forest area or volume as a result of deforestation. 

Ecosystem Any complex of living organisms together with all the other biotic and abiotic factors which 
affect them that is isolated for purposes of study, e.g. a forest ecosystem is a part of a forest 
being uniform in climate, parent materials, physiography, vegetation, soils, animals, and 

Ecotowim Nature tourism (low-impact tourism) 
Edefk Of the soil. 

Environmental services Beneficial functions performed by natural forest ecosystems, including the maintenance 
of biodiversity, protection of soil and water resources, moderation of climate, influence on 
rainfall, sequestering of carbon dioxide, provision of habitat for wildlife, and maintenance of 
the earth's natural balance. 

Externality A cost (or a benefit) of an economic activity by one party that is unintentionally imposed on 
(or received) by another party without compensation (or payment) that leads to inefficiencies 
in competitive markets. 

Facultative halophytes Plants inhabiting, but not restricted to, salty soils as opposed to Obligate halophytes, 
which will grow only where the salt levels in the soil are high. 

Felling cycle The interval between successive main fellings in the same area under the selection system. 

Felling Series A forest area forming the whole or pan of a working circle and delimited so as, (1) to 
distribute felling and regeneration to suit local conditions, and (2) maintain or create a normal 
distribution of age classes. 

Forest biamass The biomass of trees, shrubs and lesser vegetation in a forest ecosystem including their below- 
ground parts. 

Geomorphology The study og the origin, evolution and configuration of the natural features on the Earth's 

Hydromorhpic soil A soil in which the effects of poor drainage is the main factor in determining its 
morphology, giving rise to a predominance of gley colours. 

Land Use Manning The process by which decisions are made on future land uses over extended time periods, 
that are deemed to best serve the general welfare. 

Levee That part of a river floodplain closest to the river and above the main floodplain level, built 

up by the deposition of relatively coarse textured material which settles when the river 
overflows its banks. 

Littoral Of, on or near the shore. Region lying along the shore, especially land lying between the 

high and low tide levels. 

Multiple Use Spatially three somewhat different ideas are involved: (1) different uses of adjacent sub-areas 
which together form a composite multiple-use area, (2) the alternation in time of different uses 
on die same area, and (3) more than one use of an area at one time. In the first two ideas it 
is implicit that direct competition between uses is avoided by alternating spatially or in time. 
The last idea involves multiple use in the sense of simultaneous use of one space and must 
concern itself with complementary versus conflicting activities, compatible and incompatible 

Where spatially coincident uses are involved at a given time, conflicts between resource users 
will almost always occur and the concept of such forms of multiple use should be realistically 
interpreted as a dominant use with secondary uses integrated only insofar as they are 
compatible with the first. However where the idea of incompatibility relates to the economics 
of productivity maximization of single resource yield, management and multiple use can 
perhaps be validated in terms other than single-resource production efficiency. 

Nonfood forest products Tangible minor forest products, such as fruit, nuts, bushmeat, fishery. 

Photic zone (Also called Euphotic zone) The pan of the water column in which there is sufficient light for 
photosynthesis to occur. The lower limet varies from a few matres in estuarine waters, which 
typically contain a considerable amount of suspended material, to more than 100 m in clear 
oceanic waters. 

Planning The determination of the goals and objectives of an enterprise and the selection, through a 
systematic consideration of alternatives, of the policies, programs and procedures for 
achieving them. An activity devoted to clearly identifying, defining, and determining courses 
of action, before their initiation, necessary to achieve predetermined goals and objectives. 

Planning Horizon The time period which will be considered in the planning process. 

Potential evapotranspiration The maximum possible loss of soil moisture under a given climatic condition, by 
transpiration through the leaves of plants and by direct evaporation. 

Sustainable development Development that meets present needs without comprising the ability of future 
generations to meet their own needs. 

Sustainable forest management Utilization of forests (including aquatic resources in the mangroves) without 
undermining their use by present and future generations. 

Strategy A broad non-specific statement of an approach to accomplishing desired goals and objectives. 

Tree biomass The biomass of vegetation classified as trees including foliage, stump and roots. (FAO defines 
a tree as a woody plant having a main stem which, when growing under normal conditions, 
reaches a mature height of at least 7 metres) 

Wood biomass The biomass of woody vegetation such as trees and shrubs including stumps and roots 



lacre (ac) 

1 cubic metre (m 3 ) 

1 cubic foot 



1 tonne (t) 

0.405 hectare 

3S.31 cubic feet 

0.028 cubic metre 

3.625 cubic metre (stacked) 

0.0605 tonne 

0.9842 long ton 


A.A.C. Annual Allowable Cut 

C. V. Coefficient of Variation 

C.A.L Current Annual Increment (m s /ha/year) 

das Diameter above stilt roots (30 cm above the highest stilt root) 

D.B.H./dbh Diameter at breast height (1.3 m above ground level) 

ffb Fresh fruit bunch (Oil palm production) 

IUCN International Union for Conservation of Nature and Natural Resources (World Conservation 

Ln Logarithm to natural base "e" 

M.A.I. Mean annual increment (m 3 /ha/year) 

M. C. Moisture content (in percent) 

o.b. Overbark 

ppt or %* per thousand (parts per thousand) 

UNDP United Nations Development Programme 


Qrowing awareness of the protective, productive and social functions of 
tropical mangrove ecosystems has highlighted the need to conserve and manage 
them sustainably. Given their multiple-use potential, it is imperative that 
mangrove -based terrestrial and aquatic resources be managed in an integrated 
manner. This implies that no-single resource use should be maximized per se 
to the point where the sustainable potential of another resource is adversely 
affected. The traditional "management paradigm" implying that if forests are 
well managed than, ipso facto, the non-wood ecosystem components will remain 
stable, is notionally flawed. Mangrove fishery, mariculture and wildlife 
management programmes have to be structured and integrated into the overall 
policy, implementation, and control levels of an integrated resource 
management system. 

These guidelines, while promoting an integrated coastal area management 
approach to mangrove ecosystems, focus on the forest management aspects. They 
provide a broad synthesis of management systems that have been successfully 
practised in Southeast Asia, and the experiences of FAO in promoting 
sustainable forest management in Africa, the Caribbean, Central America and 
in other tropical mangrove areas. The present document looks at mangrove 
management from a broad perspective that goes beyond wood production per se. 
It is organized as follows: 

Part I focuses on the ecological and biological foundation for 
sustainable management planning within a multiple-use framework , 
including a brief review of relevant literature. 

Part II deals with the multiple-use potential of mangroves and 
discusses the utilization of selected mangrove -based products. Land 
use and protection aspects are also covered. 

Part III covers the assessment of mangrove forest resources through the 
use of remote sensing, surveying, mapping and forest inventories; 
highlighting areas where these techniques differ from conventional 
approaches due to the specific characteristics of the mangroves. 

Part IV focuses on the application of technical, managerial, economic 
and human resources to manage and use mangrove resources sustainably to 
meet the needs of people, and as a tool in rural development without 
impairing the environment. An objective assessment of the 
environmental impacts of mangrove forest management is attempted. 
Conclusions and recommendations are also presented. 

Five small case studies dealing with various aspects of mangrove 
resource assessment are included in the back of 'this document together with 
a larger case study on Multiple-use Management of the Sundarbans Forest in 

In structuring an appropriate response to manage mangrove ecosystems 
sustainably, within an integrated multiple-use context, it is necessary to 
recognize that there are as yet many information gaps and constraints. These 
per se should not be considered as impediments to initiating mangrove 
management, as much empirical knowledge can be obtained by following and 
adapting the experiences gained in other countries. 



The mangrove formation is the gift of the land and sea. Mangroves 
depend on terrestrial and tidal waters for their nourishment, and silt 
deposits from upland erosion as substrate for support. The tides nourish the 
forest/ and mineral rich river-borne sediments enrich the swamp. Thus the 
mangroves derive their form and nurture from both marine and terrestrial 

It is one of the most productive ecosystems and a natural, renewable 
resource. However, on all sides, the world's mangroves are beleaguered. 
Mangroves in particular are losing their habitats, as rivers are dammed, their 
waters diverted and the intertidal zone extensively developed for agriculture 
or aquaculture and generally dried up. Large tracts are converted to rice 
fields, industrial and land development and other non-wood uses. In response 
to the lucrative shrimp export trade, a new breed of small and large-scale 
farmers are carving out large chunks of tidal flats for shrimp farming and 
pisciculture. In parts of Asia, the mangrove is home to thousands of 
families. Mangrove areas are over-exploited for fuelwood and charcoal -making. 
In over -populated and acute fuelwood deficit areas, even small branches and 
saplings are removed primarily for domestic fuel. 

The depletion of mangroves is a cause of serious environmental and 
economic concern to many developing countries. This stems from the fact that 
at the interface between the sea and the land, mangroves play a pivotal role 
in moderating monsoonal tidal floods and in coastal protection. At the same 
time their primary production supports numerous forms of wildlife and avifauna 
as well as estuarine and near-shore fisheries. Consequently, the continuing 
degradation and depletion of this vital resource will reduce not only 
terrestrial and aquatic production and wildlife habitats, but more 
importantly, the environmental stability of coastal forests that afford 
protection to inland agricultural crops and villages will become seriously 

Habitat protection is the ultimate goal of conservation, to which all 
other approaches are subsidiary. For conservationists worldwide, mangroves 
present the great immediate challenge. 

Technically, mangroves are easier to manage compared to the species rich 
humid tropical forests. There are typically only a handful of mangrove 
species, many of which coppice or regenerate freely. However, whereas the 
terrestrial forester is concerned primarily with managing forests grown on 
stable and firm ground, in the tidal swamps he has to manage aquatic resources 
as well, on a substrate that is ever changing and dynamic over time. 

Given the multiple use potential of mangrove ecosystems and their 
linkages to terrestrial land use, an integrated approach is needed. Clearly, 
an integrated conservation and sustainable management approach, in the face 
of weak informational database prevailing in most countries, will require 
years of management research. Modern agricultural farming systems based on 
high yielding crops have evolved through centuries of concerted field 

Mangrove forest management, in comparison, is a relatively new science, 
and unlike its agricultural counterpart, deals with long term tree crops 
rather than short term food crops. The economic and social reality in most 
developing countries is simply, that forests cannot be conserved unless they 
are productively used. However, mere use of a resource is not management. 
It is important therefore to demonstrate potentially viable management 
alternatives to convince decision-makers to forestall some of the mangrove 
forest conversions and destruction until more detailed and reliable 
information is available. 

In support of the foregoing, the Forest Resources Division of the 
Forestry Department in FAO has singly, and in cooperation with other UN 
agencies, promoted seminars and workshops on mangrove management, notably in 
Central America, the Caribbean, West Africa and the Asia Pacific region. 
Through Technical Cooperation Programme arrangements, integrated forest 
management models have been developed for Cuba, Costa Rica, Vietnam and 
several other countries. FAO has also provided technical expertise in 
mangrove management to assist countries such as Panama, Guyana, Ecuador, 
Sierra Leone, Kenya, Bangladesh, Myanmar, Thailand, Indonesia and Papua New 
Guinea to name a few. 

Diverse management approaches designed to meet location-specific 
situations, priorities and needs have been tried. Thus, in Cuba, the acute 
shortage of railway sleepers has prompted the experimental use of portable 
cable winch systems, adapted to swamp working conditions. The management plan 
in this instance includes ecosystem management, honey production as well as 
improved carbonization techniques. In Costa Rica, an innovative integrated 
management plan to promote multiple use of molluscan resources, along side 
bark production and charcoal has been introduced for the very first time. 
Restoration of the degraded coastal mangrove belt to control the ingress of 
salt water constitutes a major technical support element in Guyana, where most 
of the coastal agricultural land is below sea level. . In Sierra Leone, the 
main thrust is to restore the biological diversity and productivity of overcut 
mangroves, afforestate degraded mud-flats and rehabilitate other human 
impacted coastal areas. In Bangladesh, studies are conducted in the 
Sundarbans to devise viable management alternatives in the face of the natural 
and/or man- induced decline of productive growing stocks of Sundri (Heritiera 
fomes) , coupled with an integrated multi-disciplinary approach to diversify 
and enhance the multiple -use potential of the area, including conservation of 
the endangered Royal Bengal tiger. In Myanmar, a feasibility study focusing 
on species trials to restock the degraded Ayeyarwady (Irrawaddy) delta 
mangroves is initiated and a M coppice-with-standards" system is now being 
tried in place of the selective felling system. In the Ca Mau peninsula in 
Southern Vietnam, a shrimp and tree farming model has been promoted as a 
sustainable land use option, to harmonise the competing use of tidal swamps 
for shrimp farming on the one hand, and the maintenance of ecological 
functions and fuelwood production on the other. 

Experience suggests that many degraded mangrove ecosystems can and do 
recover from intensive and sometimes catastrophic human activities such as the 
massive herbicidal destruction of the mangroves around Ho Chi Minh City and 
the Ca Mau peninsula in Southern Vietnam during the Indo-China conflict. 
Nevertheless, the fact that ecosystems recover from gross human disturbance 
does not mean that the costs and socioeconomic consequences of such impacts 
are low. 

An appreciation of key physico-biological interrelationships and 
ecological -human development interactions against the background of prevailing 
human needs is pivotal in the formulation of successful management strategies, 
consequently peoples' participation is crucial. In many situations the 
technical problems are less insurmountable than the social ones. 
Traditionally, foresters are conservative and tend to think in terms of 
management within the narrow territorial confines of reserved forests and 
applying sustained yield management to such reserves, and then only onto the 
so-called productive parts. The productive forest areas may constitute only 
a small part of the overall tidal swamp which has many ecological sites and 
niches, all of which play an important part in the overall ecological complex. 
Nowhere is the need to take a holistic environmental approach more crucial 
than in mangrove ecosystems where terrestrial, coastal and human influences 
are key interacting factors. 

Until and unless foresters are prepared to share their "reserved 
mangrove domain" with other land-users and accept that, in certain cases, non- 
foresters, who may be marine biologists or ecologists, can manage the mangrove 
ecosystem better than they do, an integrated approach will be difficult to 
apply. Indeed, one of the key institutional constraints common to most 
integrated projects is weak or lack of coordination between different land- 
users and concerned agencies. One way to alleviate this problem is by 
creating a National Mangrove Committee, comprising concerned ministries and 
departments, research and education institutions and NGOs and charge it with 
the development and implementation of integrated mangrove management plans 
and/or the monitoring and evaluation of these. In some countries an agency 
responsible for integrated coastal area management (ICAM or ICZM) may already 
be in place, which could perform the above task given additional assistance 
from mangrove specialists. 

Where the concern is environmental, the data needs must not only be 
comprehensive enough to cover all aspects, but the manner in which the data 
is collected and analyzed has to be undertaken in an integrated way using a 
mult i -disciplinary approach with the identification and characterization of 
key inter-relationships in mind. A useful method for the design of integrated 
resource management plans and sustainable land use is to conduct extensive 
resource and land use surveys supported by selected ecological and social 
studies. The people are the key actors in any plan. Their support and 
participation from the very beginning are crucial in implementing a successful 
participatory forestry programme. 

These guidelines, while promoting an integrated coastal area management 
approach to mangrove ecosystems, focus on the forest management aspects. They 
provide a broad synthesis of selected approaches applied, experiences gained 
and lessons learned. However, the present document looks at mangrove 
management from a broad perspective that goes beyond wood production per se. 
The aim of this wider perspective is to highlight and promote management 
approaches to environmentally sustainable, multiple-use management through 
commitment towards rational land use and greater responsiveness to people's 
concerns and needs. This does not imply that classical management paradigms 
are unsound but rather that these should be modified as appropriate in 
response to the heightened expectations of people and the imperatives for 
sustainable environmental management that transcend sustainable timber 
management . 

The guidelines are organized as follows: 

Part I focuses on the ecological and biological foundation for 
sustainable management planning within a multiple-use framework, 
including a brief review of relevant literature. 

Part II deals with the multiple-use potential of mangroves and 
discusses the utilization of selected mangrove -based products. Land 
use and protection aspects are also covered. 

Part III covers the assessment of mangrove forest resources through the 
use of remote sensing, surveying, mapping and forest inventories; 
highlighting areas where these techniques differ from conventional 
approaches due to the specific characteristics of the mangroves. 

Part IV focuses on the application of technical, managerial, economic 
and human resources to manage and use mangrove resources sustainably to 
meet the needs of people, and as a tool in rural development without 
impairing the environment. An objective assessment of the 
environmental impacts of mangrove forest management is attempted. 
Conclusions and recommendations are also presented. 

Five small case studies dealing with various aspects of mangrove 
resource assessment are included in the back of this document together with 
a larger case study on Multiple-use Management of the Sundarbans Forest in 

For additional case studies on mangrove management and utilization the 
reader is referred to two previously published papers by the FAO dealing with 
Asia and the Pacific (FAO Environment Papers 3 and 4) and the FAO 
Miscellaneous Paper (FO:MlSC/86/4) 'Ordenacion Integrada de los Manglares' by 
B.Rollet which covers Latin America. ISME/ITTO (1993a) also present a summary 
of the status of mangrove management in Latin American countries. With regard 
to Africa only limited studies are available, and the reader is referred to 
SECA/CML (1987) and ISME/ITTO (1993b) for initial information. 

For further information on integrated costal area management refer to 
FAO Fisheries Technical Paper 327 (FAO, 1992) and Chue et al. (1991). A 
series of integrated management plans for coastal zones in South East Asia 
have also been published by ICLARM (International Center for Living Aquatic 
Resources) as part of the ASEAN/US Coastal Resources Management Project. A 
number of these zones include mangrove areas. 



This chapter reviews some ecological and biological relationships of 
mangrove ecosystems, particularly in respect of those biophysical aspects 
relevant in designing appropriate silvicultural systems and in sustainable, 
multiple-use management planning. A brief review of relevant literature is 
also presented. 



Mangroves are the characteristic littoral plant formations of tropical 
and subtropical sheltered coastlines. They have been variously described as 
' coastal woodland', 'tidal forest' and 'mangrove forest' 

Generally mangroves are trees and bushes growing below the high-water 
level of spring tides (FAO, 1952) . Their root systems are thus regularly 
inundated with saline water, even though it may be diluted due to freshwater 
surface run-off s and only flooded once or twice a year. 

Macnae (1968) suggested that the word 'mangrove' should be used for the 
individual trees and bushes, whereas 'mangal' be referred to communities of 
such plants. This usage has not been followed here because the context 
usually makes it clear whether one is referring to 'mangrove trees' or a 
'mangrove forest' . 

2.1.1 Floristics 

The most noteworthy features of the mangrove forests, apart from 
their unique habitat, are the relative paucity of the species 
comprising them; the arch- formed stilt roots of the Rhizophora spp; the 
clusters of blind root suckers or pneumatophores from other genera such 
as Avicennia and Sonneratia, which protrude from the ground in such 
numbers as to be an impediment to walking; the curious adaptation to 
environment by which the seeds of the Rhizophora germinate on the 
trees; as well as the high incidence of trees with lenticellated bark. 

The mangrove forests are evergreen. The paucity of species 
occurring in them is due to the peculiar conditions of their existence, 
few plants being able to tolerate and flourish in saline mud and to 
withstand frequent inundation by sea-water. They also differ from the 
inland forests in that certain species are practically gregarious over 
extensive areas. Apart from the Rhizophora spp, many of the principal 
species coppice readily. The flora that comprises arborescent species 
with forestry importance is confined to a few families, viz., 
Rhizophoraceae, Combretaceae, Avicenniaceae/Verbenaceae, Meliaceae, 
Sonneratiaceae, Sterculiaceae, Buphorbiaceae, Theaceae and 
Pelliceriaceae. Other families are sparingly represented, chiefly in 
situations where the limits of the mangrove are not sharply defined. 

In the supra- littoral and inter- terrestrial zone, where brackish 
water conditions prevail, there are species that occur in the mangrove 
habitat proper, but which are not restricted to it, e.g. Acrostichum 
aureum, A. speciosum and A. danae folium. Recognizing this, Saenger 
et al. (1983), has divided the mangroves into two broad groups 
comprising the exclusive species that are restricted to the mangrove 
habitat, and the non-exclusive species, which may be important in the 
mangrove habitat, but are not restricted to it. 

Species normally found in seasonal swamps that are subject to 
occasional saline influences, are dealt with superficially in this 
report, as their silviculture and management requirements are 
distinctly different (e.g. Mora oleifera, Annona grlaJbra, and 
Pterocarpus officinalis) . 

2.1.2 Mangrove taxonomy 

Depending on the concept of mangroves and definition of mangrove 
habitat applied, the number of species cited by different sources 
varies significantly (Tabl* 2.1). This points to the need to 
standardize the criteria used in the definition and delimitation of 
mangrove habitats (Jimenez and Soto, 1985). 

Table 2.1: Number of mangrove species 

Source Family Qanara Spaciaa 

Source: I/ based on Chapman (1970, 1974) 

Lugo t Snadafcar, (1974) 

Saangar at al, (1983) 

Cintron & Schaaffar-Novalli 

Chapman (1970)/Walah (1974) 

Chapman, (1974) 

Blaaco, (1984) 

Narcar t Hamilton, (1984) 

In Table 2.2 on the following page, most trees and shrubs of the world's 
mangroves are listed after Saenger et al. (1983) and in Figures 2.1 - 2.4 
representatives of some of the most common genera are illustrated. For 
additional information on the individual mangrove species and keys to their 
identification, the reader is referred to Chapman (1976); Tomlinson (1986); 
Watson (1928) and a recent field guide by Aksornkoae et al. (1992) . 


The general distribution of mangroves corresponds to that of the 
tropical forests, but they extend further North and South of the equator, 
sometimes beyond the tropics, although in a reduced form. 

Floristically, two main zones can be distinguised : An Eastern zone 
comprising the Bast African coast, South Asia and the Pacific including the 
islands down to Australia, and a Western zone including West Africa, the 
coasts of America and the Carribean (PAD, 1952) . 

Table 2.2: A list of mangrove spedes world wide 

A-Eclusive species 


B-Some important, nonexclusive species UMbon 

Acanthus ebracteatus Vahl. S 

Acanthus ilictfbtius L. S 

Acanthus volubUis Wall. S 

Aegialitis anmttota R.Br. S 

Aegialitis rotundtfolia Roxb. S 

Aegiceras comicutotum (L.) Blanco S 

^ vicennia alba Blume T 

/4 vicennia bicolor Standl. T 

A vicennia eucatyptifolia Zipp. ex Miq. T 

Avicennia germinans L. T 

Avicennia intermedia Griff. T 

Avicennia lanata Ridley T 

Avicennia marina (Ponk.) Vierb. T 

Avicennia officinalis L. T 

Avicennia rumphiana Hall.f. T 

X vicennia tomentosa Willd. T 

Avicennia tonduzii Moldcnke T 

Bruguiera cylindrica (L.) Blume T 

Bruguiera exaristata Ding Hou T 

Bruguiera gymnorhiza (L.) Lam. T 

Bruguiera hainesii C.G.Rogers T 

Bruguiera parviflora (Roxb.) Wight & Am. T 

Bruguiera sexangula (Lour.) Poiret T 

Camptostemon philippinensis Becc. T 

Camptostemon schultzii Mast. T 

Ceriops decandra (Griff.) Ding Hou T 

Ceriops tagal (Perrottet) C.B.Robinson T 

Conocarpus erectus L. T 

Cynometra iripa Kostel T 

Cynometra ramiflora L. T 

Excoecaria agallocha L. T 

Heriticra littoralis Alton ex Dryander T 

Heritiera fames Buch.-Ham. T 

Kandelia candel (L.) Dnice T 

Laguncularia racemosa Gaeitn.f. T 

Lumnitzera littorea (Jack) Voigt S/T 

Lumnitzera racemosa Willd. S/T 

JVypa fruticans van Wurmb. P 

Osbornia octodonta F. Muell. S 

Pelliciera rhizophorae Planchon & Triana T 

Phoenix paludosa Roxb. P 

Rhizophora apiculata Blume T 

Rhizophora harrisonii Leechman T 

Rhizophora x larmardni Montrouz T 

Rhizophora mangle L. T 

Rhizophora mucronata Lam. T 

Rhizophora racemosa G.Meyer T 

Rhizophora x selala (Salvoza) Tomlinson T 

Rhizophora stylosa Griff. T 

Scyphiphora hydrophyllacea Gaeitn. S 

Sonneratia alba J. Smith T 

Sonneratia apetala Buch.-Ham. T 

Sonneratia casedaris (L.) Engl. T 

Sonneratia griffiM Kurz T 

Sonneratia ovata Backer T 

Xylocarpus australasicus Ridley T 

Xylocarpus gangeticus Parkison T 

Xylocarpus granatum Koenig T 

.Xylocarpus moluccensis (Lam.) Roem. T 

Xylocarpus parvtfolius Ridky T 

Acrostichum aureum L. 
Acrostidutm danaefolium Langid. & Pish 
Acrostichum spedosum Willd. 
Barringtonia racemosa Roxb. 
Brountonw argentata Kurz. 
BnmfiloM^i /frw (L.) Kosterm. 
Cerberafloribunda K.Schum. 
Cerbera manghas L. 
Clerodendrum inerme (L.) Gaeitn. 
Cynometra mannU Oliver 
Dimorphandra oUrfera (Triana ex. Hemsl.) 
Dolichandrone spathacea (L.F.) K.Schum, 
Mftfccitt /tomote Sieb & Zucc. 

Mauritia flexuosa (Linn.f.) 

Maytcnus emarginata (Willd.) Ding Hou 


OncospermaJUamentosa Bl. 

Pemphis addula Porster 

Pterocarpus officinalis Jacq. 

Thespesia acutiloba (E.G.Backer)Excell & 


Thespesia populnea (L.) Soland. ex Coir. 
Thespesia populneoides (Roxb.) Kostel 

T * Tree 
S = Shrub 
P = Palm 
F = Fern 





Source: Saenger el al. (1983) 



All the genera (but not all the species) of the Western mangroves are 
found in the Eastern zone, but the latter is far richer in number of different 
species. Numerically! there is a five- to-one disparity in species numbers in 
the two groups. Chapman (1975) theorizes that the oceanic current around the 
Cape of Good Hope prevents the drift and migration of species from the Indo- 
West-Paeific biogeographic region to West Africa and the Atlantic. A 
generalized global distribution of mangroves and the number of species in 
various regions is shown in Figurs 2.5 



Weftetn. roingfovei 



figure 2.5: Generalized dbtributton of mangroves 

The total area of mangroves in the world is not well known. In 
Table 2.3 approximate mangrove areas in various countries are represented. 
More recent data on the extent of the mangrove formations in some of the 
countries in Latin America and Africa have been reported by ISME/ITTO (I993a, 
1993b) . However, as the accuracy of the estimates varies widely from one 
country to another, the need for an updated, world wide survey is evident. 

The world's greatest contiguous mangrove area is the Sundarbans situated 
in the Bay of Bengal, which covers a total land area of approx. 660 000 ha. 

Due to their situation along coastal lines, mangrove formations are 
constantly controlled by marine and terrestial factors such as local climate, 
geomorphology, salinity and other edaphic characteristics. These, together 
with the distance from the sea, the frequency and duration of inundation and 
tidal dynamics, govern to a great extent the local distribution of species and 
their succession* 


Table 23: Approximate mangrove areas in various countries 


Area (Ha) 


Area (Ha)** 


Area (Ha)*** 


















307,000 1 





Costa Rica 

39,000 1 






448,000 I 



Guinea Bissau 


Dominican Rep. 

9,000 I 





El Salvador 







196,000 1 





French Guiana 

55,000 I 




few ha. 


3,000 1 

Papua New Guinea 553,000 
Philippines 240,000 





50,000 1 
150,000 1 

Sri Lanka 





18,000 I 



Sierra Leone 



145,000 1 























Trinidad and 


4,000 1 

USA (Florida 

+ P. Rico) 

178,000 1 







Sources: (+) 




Saenger et al 

. (1983) 



FAO (1981, 1986) 1 

Best developments of the mangroves are found at locations with deep, 
well aerated soils, rich in organic matter and low in sand, usually in 
estuaries. Christensen (1983) indicated that JUiizophora may attain heights 
of more than 40 m in such areas. 

The inland extend of mangroves depend on the morphology of the soil and 
the factors mentioned above, and the forests can vary in size from a few 
clusters of small trees or shrubs to extensive areas of well developed stands. 


"Ecology is the science of the interrelationships of organisms in and 
to their complete environment". (Spurr and Barnes, 198.0). For forest 
communities, the names of the predominant tree species which make up the 
characteristic physiognomy of any given stand are used to classify the forests 
into "forest types". Thus a "JRhizpphora forest type" is characterised by the 
predominance of Rhizophora species, conjuring mental images of trees typified 
by prop or aerial roots as well as elongated pendulous propagules. 

The forest community and its habitat constitute an ecosystem, in which 
the constituent organisms and their environments interact in complex processes 
and life cycles of carbon, water, and nutrients. Studies of the forest 
ecosystem takes into account both the organic and inorganic aspects of the 
cyclic processes of life, and are increasingly given greater focus due to the 
growing need for environmentally sound forest management. 

The physical forest ecosystem environment and its biotic factors 
constitute the habitat or site. The forest environment is the composite end 
product of many interacting forces, hence site is the sum total of 
environmental conditions operating at any one place. 


The ultimate goal of forest management, economic considerations aside, 
is to exploit to the fullest the natural energies and resources available for 
any given site so as to produce maximum carrying capacity for the production 
of the desired products. A careful examination of forest site conditions will 
prove to be a worth while investment both in time and effort for any forester, 
in so far as it portrays the potential stand productivity under ideal 

Forest ecosystem analysis requires multidisciplinary investigations for 
a comprehensive understanding of system dynamics. The purpose of all these 
studies is to better understand and to predict changes that are likely to 
occur when ecosystems are subject to stresses and manipulations. However, in 
practice, a total analysis is seldom possible given the human and financial 
resources available. Forest managers should therefore be aware of the 
limitations under which they work and partly because of this awareness, 
conventional wisdom dictates that natural forests should be managed 
conservatively rather than rigidly according to economic goals. However, a 
gravitation towards the systems approach, which involves an assessment of 
systems productivity and their underlying process functions, is inevitable 
given the fact that increasingly forest managers are expected to resolve 
environmental issues. Systems -oriented studies allow us to perceive and 
resolve environmental problems in a different way. As stated by Reichle 
(1971) : 

Geologists cannot continue to respond to each new environmental crisis by 
simplistic 'cause and effect " studies of isolated ecosystem components. The totality 
of environmental systems must be recognized and an understanding developed of the 
interactions and interdependences of systems components. Only in this vtwy can 
the effects of perturbation upon individual components be interpreted in the total 
context of the system*. Reichle (1971). 

In the following the main abiotic factors influencing the mangrove 
ecosystem, viz. climate and edaphic factors, are described followed by a more 
detailed description of the main biotic elements constituting the ecosystem 
i.e. mangrove flora and fauna, their inter-relationships and the management 
implications of these. 



Pannier and Pannier (1977) broadly summarized the present knowledge 
concerning the distribution of mangrove forests in relation to climatic 
regions. According to Walter (1977), mangrove ecosystems are mainly found in 
three climatic divisions, viz., (a) the equatorial zone, between approximately 
10N and 5-10S; (b) the tropical summer-rainfall zone, north and south of the 
equatorial zone, to approximately 25-30N and S, partly in subtropical dry 
zone of the deserts, still further poleward; and (c) partly in warm temperate 
climates that do not have really cold winters, and only on the eastern border 
of the continents in this zone. 

Blasco (1984) suggests that both temperature and rainfall should be 
shown in a single climatic diagram, because they are essential bioclimatic 
factors for mangroves and other terrestrial plants. 


The length of the rainy season determines the influx of freshwater in 
each site. In equatorial climates, the upland runoff is usually adequate to 
maintain freshwater in contact with the saline water table throughout the 
year. In dry monsoonal climates, salinity in the upper soil layers increases 
during the dry season. The number, duration and intensity of dry seasons, 
therefore, directly influence the distribution of salinity in the intertidal 
zone. The alteration of upland drainage may have an impact on those mangroves 
requiring recharge of freshwater. 

Shoot growth of seedling and sapling trees is closely correlated to 
water potential within the plant and to environmental soil moisture deficits. 
In favourable sites, shoot growth occurs in varying degrees throughout the 
year but in high- stress sites, shoot die-back occurs during prolonged dry 
seasons and vigourous growth occurs mainly during the rainy season. 

2.4.1 Temperature 

In the equatorial belt, temperature is usually not a constraining 
factor with regard to plant growth. However, periods of intense 
physiological stress may be experienced when high temperatures are 
combined with full sunlight and prevailing winds giving rise to high 
evapotranspiration and increased surface salinity due to capillary 
uptake. In such cases the formation of heavy salty crusts on the soil 
surface can be harmful to plant growth. In the Guanal area in Cuba, 
"salitrales" are common in the more exposed coastal sites. (Chong, 
1989b) . Salt-flats (albinas) also occur naturally in the Aguadulce 
area in Panama. 

2.4.2 Winds and storms 

The impact of severe storms on the forest can be profound. In 
areas that are exposed to severe storms the canopy of the forests along 
the coasts is usually broken. Structurally, the trees are also 
shorter. This partly explains the fact that high mangroves are found 
generally in more sheltered situations. 

Severe storms (cyclones/hurricanes) affect the waves, swell, 
storm tides, and current system, as well as the volume and rate of 
fresh water discharge from the land. (Riggs, 1977) . Mangroves play an 
important role in moderating coastal storms at the interface between 
the land and sea. The coastal belt, particularly near and along the 
foreshore, is a zone of intense atmospheric turbulence due to the 
interplay of land and oceanic atmospheric influences. The impact of 
cyclones on densely populated deltas can be tragic. In November, 1970 
a cyclone, combined with high tide, killed more than 200 000 people in 
Bangladesh. The 1991 cyclone reportedly killed over 100 000 people 
and rescue work was hampered by high tidal floods. Without the 
moderating influence of the forest, the loss of human life and property 
is catastrophic. Environmentally, therefore, the coastal mangroves are 
vital shelter-belts which afford protection to inland homesteads, 
agricultural crops, livestock and aquaculture. 

Along the coasts, a belt of protective mangrove vegetation should 
always be retained not only to reduce the damaging effects of tidal 
waves and storms but also to reduce the severity of tidal flooding. 


2.4.3 Rainfall 

Mangroves do not rely absolutely on rainfall for survival because 
they can extract fresh water from the sea through salt excreting 
glands. (Chapman, 1976) . However, the amount of rainfall influences 
mangroves in two ways: (1) as rainfall determines the rate of 
weathering it accounts for the amount of silt brought to the mangrove 
swamp, and (2) high rainfall reduces the incidence of hyper -salinity. 
According to Macnae (1966, 1968), Australian mangroves thrive best in 
areas receiving more than 2 500 mm of rain per year, as salt flats are 
often formed in areas with a precipitation of less than 1 500 mm/year. 

Where the mean potential evapotranspiration (PET) is high and the 
amount of rainfall is insufficient to reduce the accumulation of salt, 
salt flats becomes predominant as is the case in Cuba, parts of Panama, 
West Africa and West India. 

2.4.4 Life Zones (Zonas de Vida) 

In Central and South America, Holdridge's Life Zone system for 
classifying of vegetation and climate has been applied to the 
compilation of ecological maps. The ecological map for Costa Rica 
prepared by Tosi (1966) at a scale of 1:750 000 provides a good 
classification even for the coastal mangroves (Holdridge 1947, 1971) . 

According to the Life Zone system, the most suitable life zone 
for silvicultural management is the Tropical Wet Forest (bmh-T) , 
followed by Tropical Moist Forest (bh-T) , and the least suitable is the 
Tropical Dry Forest (bs-T) . Forests lying outside these life zones have 
strong silvicultural limitations. 

Chong (1988) observed a similar correlation between the life zone 
types and forest productivity classes in the Terr aba -Sierpe mangrove 
area in Costa Rica. However, given the multiple-use potential of even 
the least productive forest types, their potential uses cannot be 
easily discounted. 

2.5.1 Mangrove creomorDholocrv 
Delta formation 

Large mangrove formations are typically found on relatively 
sheltered deltaic littoral plains. In Bangladesh, the Ganges and 
Brahmaputra monsoonal waters overflow the river banks almost every 
year, depositing sediments over alluvial flats, riverine and tidal 
plains gradually extending the delta Southwards into the shallow waters 
of the Bay of Bengal. 

Delta formation is a delicate balance between the type and amount 
of river sediment, compatibility of the sediment, vegetation, changes 
in sea levels, the underlying geology and geomorphology, and the wave 
and tidal forces where the river meets the sea. 


Subsidence of sediments is a common phenomenon. Rivers carrying 
muddy sediments that compact much more that sand or silt produce deltas 
that are more prone to subsidence. When this clay settles, as much as 
80 per cent of its initial volume is water, which is gradually squeezed 
out by new sediments deposited on top. This causes the land of the 
delta to gradually subside unless new superficial mud is regularly 
deposited. Due to natural subsidence, the central portion of most 
tidal swamps, which is not well flushed by tides, tends to be more 
prone to deep flooding. However, where the deposits are calcareous, 
the substrate is more compact and consolidated, and the rate of 
subsidence is slower (e.g. in the Sundarbans and the Ayeyarwady deltas 
in Bangladesh and the Union of Myanmar respectively) . 

When river silt load is reduced by human interference, such as 
river diversion, dams or channels dug to facilitate navigation, the 
dynamic balance between land erosion and accretion will be adversely 
affected, leading to land subsidence in many areas. The Aswan High 
Dam, acting as a silt trap, has effectively halted the Nile delta- 
building process and initiated an active process of coastal retreat due 
to erosion and subsidence. (Kassas, 1972) . Farmers that settle along 
rivers often build levees that can impede water movement thereby 
disrupting the natural siltation process. 

Even without human intervention, delta building is not uniform. 
Rivers change their courses, to seek shorter, steeper paths to the 
ocean. Little or no sediment then accumulates along the abandoned 
channel, and the land around it subsides. The river's sediments are 
carried along the new channel . The delta* thus goes through a natural 
cycle of growth and decay (See Box 2.1 on the following page.) 

Subsidence is also assisted by a rise in sea levels due to 
geological processes. In the gangetic plains a general westward tilt 
in the landmass has been advanced as a cause for the eastward drift of 
rivers and accounts for the increasing salinity levels in the western 
part of the Sundarbans due to the reduction in fresh water flow. 

However, the scale Table 2.4: Global wanning scenarios 

of change associated with 
these localized geological 
phenomena is very small 
compared to the ecological 
and physical effects on 
water levels due to global 
warming: experts generally 
agree that temperature and 
sea level changes of the 
magnitudes as shown in 
Table 2.4. to the left are 
entirely probable. 

The socioeconomic and ecological costs to countries with 
substantial low lying coastal plains and populated tidal swamps can be 
very high. In some of the Pacific atoll countries, Bangladesh and 
Guyana, the social cost in the loss of agriculture land due to salt 
water intrusion, increased flooding and depletion of productive 
mangroves could be very high indeed. 

Climate effect of emission scenarios * Year 2090 


Temperature rise 

Source; Second Wortd Climate Conference, Geneva, 1990 

Business as -usual 
Scenario B 
Scenario C 


The Ddta taBdtag proem 

Delta are tte tod pro**** of erosfon. Rate washes soil tact weatbertd rock fragments Into 
streams a^ rivtri towd* the sem, Most oftbe debris it deposited along the way. where it 
fora* sediments, tod evertaatty sstt&aatery rooks. Deposits formed it the river tnouftk are 
known as die ddta, named alter the Greek tetter A, because the ancient Greek historian 
Herodotus noted that the Nik delta in Egypt, was similarly shaped. 

How far the debris travels deports 00 bow heavy it is, and how fa* the river flow*. The lighter 
the grim* of saad* ailtaad mud, and the itrongcr the current, die farther the river can wry it. 
Particles settk aa the river slows. Deltas form as the river water spreads out and slows down 
where it flows into the sea. 

Large boulders and rocks, are tardy transported very far downstream, except during catastrophic 
floods, they generally form broad, sloping "alluvial fans" that- can spread out onto lowlands at 
the edge of a range of hills. 

The form of the sfitfharrti is depeodant on the parent material of the eroded substrate, whether 
it is sandy (arenaceous) or clayey (argillaceous). Similarly, the chemical nature of the debris will 
minor the chemfoil composition of the eroded material, such as calcareous limestone. 

Smaller particles, such as sand and silt, settle as the river slows down while it travels over 
plains. These sffdimrnts build plains lying close to aea level, such as those that make up 
the Mekong and Ayeyarwady deltas. 

Sediment settles in a river's channel, especially where it meanders, and eventually blocks it. This 
forces the river to seek a new path, unless the silt is removed by dredging. When rainwater 
ova-flow* its banks after heavy monsoon rains, silt and mud spread over the surrounding flood 
plains. Such flood* generally cany more material than normal river flow adding sediments across 
the land when they recede. Flooding brings fertile silt to the plains and tidal swamps, thereby 
wricking the mangrove ecosystem. 

are not deposited evenly. Sometimes, the river deposits more debris along its banks, 
where It floods most frequeotly. This builds up "natural levees" along the edges oftbe river, 
mskrag these banks higher and drier than the surrounding land. Confined between these natural 
the river rises as silt settles, and am become higher than the surrounding land. 

The size and ihape of a delta depends on the sediment load* river flow rate and the wave power 
as wen as tidal range of the ocean. These combination of Actors enable geomorpbotogist to 
classify deltas as 'river, wave or tide" <tommafrd. Where the waves and tides are weak, the 
coast can be irregular; convoluted m shape, e.g. the Mississippi. This is because the river baa 
carried more sodimnrtfi than the tides and current can cany, so the delta grows out to the sea. 
Where waves are powerful such as at the mouths of the Nile, Ayeyarwady and Senegal Rivers, 
(be coast line is smoothly curved. Differences in wave energy can be enormous, for example 
it takes the Senegal coast only a little over 2 hours to receive as much wave power as the 
Mississippi receives all year round. Large tidal amplitudes, like strong waves, tend to smooth 
the owstSae, giving a defta the staple triangular shape. 

BOX 2.1: The Defta building process 

The restoration of protective mangrove belts must now be given 
prominence in areas such as these. At the interface between the land 
and sea, mangroves form the first line of defense against the elements 
and thus will have an increasingly important role to play if the 
adverse scenarios in terms of human suffering and economic costs 
associated with the predicted rise in sea levels are to be averted. 

Hvdro locrv arid drainage 

The quantity of freshwater discharge into the mangroves depends 
on the size of the watershed, the climate, river flow characteristics 
and diversion of water for other land uses. Where the flow is highly 
seasonal, extensive flooding usually occur during the monsoon months, 
especially when peak flows coincide with high spring tides. 

Snedaker et al., 1977, have summarized the role of fresh water from the 
Ganges through its distributaries' discharge into the Sundarbans 
estuaries as follows: 

as a salt water dilutant; 

in protecting fry, shrimp and shell -fish and other biota; 
in modifying water temperature; 
in osmoregulation of marine animals; 

as a vehicle for primary nutrients and metabolic waste removal; 
as a moderator of concent rat ion- dependent reactions in salt 

as a resource-partitioning mechanism in coastal waters; 
in the vertical movement and distribution of organisms; 
as a cutting and filling mechanism; 
in maintaining a salt wedge and mixing zone; 

in the delivery of allochthonous materials to estuaries as a 
function of precipitation, drainage and topography; 

as related to times of arrival and departure of migrating 
species . 

Sediment load and turbidity 

The amount of suspended matter transported by a river is 
dependent on its velocity. The greater the rate of flow or current, 
the larger the bed load and particle size carrying capacity. This has 
many implications for any given mangrove ecosystem, in that it affects 
the estuarine depositional pattern as well as the biology of many of 
the organisms living in the aquatic and terrestrial systems. 

The discharge of large volumes of colloidal mud from the Amazon 
river for instance makes even the establishment of hardy mangrove 
species such as Avicennia spp very difficult as the mud tends to clod 
up the lenticels in the pneumatophores . This body of mud is carried by 
the current all the way to French Guiana and Surinam, and is commonly 
referred to as the "mud sling 11 . 

Coastal -estuarine classification 

There are several definitions of 'estuary' and 'coastal area' in 
relation to mangrove formations. Many are based on geographical or 
geomorphologic descriptions. Estuaries are dynamic in behaviour, in 
that their boundaries exhibit temporal and spatial fluctuations 
reflecting local changes in river flow, wind stress, wave- tidal 
dynamics, or far- field forcing such as the synoptic wind stress on the 
coastal ocean surface and continental shelf waves. Thus the dynamic 
definition proposed by Kjerfve (1984) is useful in ecosystem management 
where an understanding of habitat dynamics, salinities and other 
factors is required. 


Kjerfve recognizes three zones as follows: (a) the estuarine 
riverine zone; (b) the estuarine mixing zone; and (c) the coastal 
boundary layer zone- An idealized estuary - coastal system is shown in 
Figure 2.6. 

IVM tt*> 




Fig. 2.6: Schematic representation of an idealized estuary - coastal lagoon (after Kjerfve) 

The estuarine mixing zone is commonly referred to as the estuary 
proper. Kjerfve sets the upriver limit at the Ifc isohaline, which 
fluctuates up or down river depending on river flow and tidal 
influences. In forest management, the lOfc isohaline is significant 
because the ferns Acrostichim sp. can become a major threat to 
regeneration in areas where the salinity is less than lOfc. This is for 
instance the case in some areas in the Matang mangrove reserve, in 
Peninsular Malaysia. It also represents the lower salinity limit for 
the white prawn Penaeus vannamei, which is the most commonly cultured 
species (Kapetsky, J.M. 1986) . 

The ebb tidal delta or river mouth bar forms the seaward 
boundary. Here the prevailing salinity is similar to that in the ocean 
waters (35fc) . The flood tidal delta and associated tidal flats, where 
they exist, are part of the estuarine mixing zone and located inside 
the geographical entrance. Typically the estuarine mixing zone shows 
marked salinity, gradients and has a very intense turbidity zone in 
which fipe grain sediments are suspended. This is the zone where 
gravitational circulation may exist, and where reversing tidal current 
tfceurs. Depending on the geomorphology, the estuarine mixing zone is 
sometime referred to as a lagoon in shallow systems with elliptical 
configurations as in the Costa del Sur mangroves of Cuba, as a fjord in 
glacially formed systems or sunken rivers as in the case of Sierra 
Leone (Chong, 1986) . 


The estuarine riverine zone is the freshwater region of the 
coastal system that experiences periodic tidal rise and fall in river 
level. In the lower regions of this zone, near the Ifc isohaline, the 
current typically reverses direction according to tidal influences. 
This zone can be very extensive as in the case of the Amazon. Fauna 
requiring substantial amounts of freshwater or which can tolerate only 
low salinities may be reared in this zone, as for example crocodiles* 

The coastal boundary laver zone is the active interface between 
the nearshore and offshore coastal oceans. It is usually characterized 
by high turbidity, high nutrient concentrations and weak salinity 
gradients. The width varies from about 1 km seawards from the river's 
mouth to more than 30 km in the case of a river or estuarine plume. 
Depending on the amount of river discharge, the configuration of this 
boundary varies and it represents the seaward extent of tidal 
influences . 

Accretion and erosion 

The mangrove habitat is a dynamic ecosystem. For optimal 
development mangroves require freshwater influence and adequate tidal 
flushing, consequently, the better stands are situated along waterways. 
In addition they are associated with a highly dynamic process of 
continual erosion and accretion which is essential to their existence 
as the continual tendency for mangrove streams to erode on one side and 
accrete on the other keeps the overall level of the formation within 
the altitudinal range of the mangrove association. 

In the case of prograding shorelines, (e.g. Matang in Peninsular 
Malaysia and in the Bay of Bengal) where extensive mud- flats are 
accreting seawards, new coastal mud- flats are formed annually. Over 
80 000 acres of newly consolidated mud-flats have been planted with 
Sonneratia apetala and Avicennia ap in the Bay of Bengal. In Matang, 
substantial areas of new mud- flats (new forests) are recorded during 
each 10 -year working plan period. 

Not all mangrove systems have prograding shorelines however. 
Three physiographic states occur in mangroves, viz., accretion, erosion 
and steady-state varying in scale. Accumulation of sediment influx or 
accretion may lead to changes in the superficial extent of a site 
and/or changes in levels, which could alter water-movement. 

These changes are particularly noticeable near the river mouths 
where water turbulence or reduced velocity of flow lead to a 
corresponding decrease in the load carrying capacity of the river and 
hence depositional behaviour. Erosion may lead to site destruction by 
scouring. In the Ca Mau peninsula in Southern Vietnam, the Eastern 
shorelines are eroding at a rate of some 300 m/year, whereas along the 
Western flank, new mud-flats are accreting at a rate of 100 m/year. 
The shoreline sedimentary pattern shows a south-westerly drift 
conforming to the general direction of coastal current and wind during 
the northeast monsoon. The third state represents a stable condition 
in which the deposition and removal of silt and mud appear to be in 
balance. This appears to be the position in the Ayeyarwady delta area 
in the Itaion of Myanmar. 


A distinction, however, should be made between natural deposition 
and erosional processes and those induced by man. Excessive silt load 
and frequent flooding due to upstream erosion and denuded river banks 
may lead to the "drowning" of mangrove roots. Similarly, the raising 
of levee banks by farmers for shrimp farming or swamp paddy often 
impedes the natural movement of water into the interior of swamps 
leading to a decrease in nutrients. Erosive waves and turbulence 
produced by motorized boats can also damage river banks, as will the 
indiscriminate removal of vegetation along streams and rivers. The 
riverine forests not only protect the river banks but are usually the 
prime source of propagules, and their removal will therefore adversely 
affect propagule supply and dispersal. 

Tides and current 

Water movement is very important to the survival of mangroves, in 
that nutrients are brought into the system by tides and from upstream 
flows. Tides carry the remains of these nutrients and dissolved 
detritus from the mangrove ecosystem further downstream to the 
estuarine systems (Dwivedi, S.N. et al., 1974; Lugo, A.E. et al. 1973). 
Water transports dissolved oxygen to the root systems of plants and 
recycles nutrients in the ecosystem (Clough, B.F. and Attiwill, P.M. 
1974) . Tides remove accumulated carbon dioxide, sulphurous toxic 
wastes, organic debris, and maintain soil salinity levels. The 
dispersal, distribution and successful establishment of propagules are 
also partly influenced by tides (Chapman, 1976, Rabinowitz, 1978) . 

Tides regulate benthonic activity. Filter feeders, such as 
clams, mussels, oysters (Mollusca) depend on the tides. Gocke, K. 
et al., (1981) have shown that tidal range and duration of immersion 
affect the relative percent of oxygen consumption of benthic organisms 
in different habitats along the Pacific coast. 

Tide is a periodic rising and falling of sea level caused by the 
gravitational attraction between the Moon, the Sun, and other 
astronomical bodies acting on the rotating Earth. The vertical rise 
and fall is called tide or astronomical tide; the horizontal movements 
of water are termed tidal current. Tides follow the moon more closely 
that they do the sun. 

Along the Pacific coast of Costa Rica, the lunar day is about 42 
minutes longer than the solar day, hence tides occur about 42 minutes 
later each day. The two daily tides on the Atlantic coast are of 
nearly the same height, but along the Pacific coast the tides have a 
pronounced diurnal inequality. The coasts of Thailand are influenced 
by 3 tidal regimes; viz., semi-diurnal along the Andaman coasts, 
diurnal tides along the Northern part of the West coasts of the Gulf, 
and the remaining coastline receives mixed types of tides, but 
prevailingly diurnal (Figure 2.3) . The Dat Mui mangroves in Southern 
Vietnam are characterized by two tidal regimes, originating from the 
Eastern coast and the Gulf of Thailand. The former is diurnal with a 
tidal range of 1-2 m, while the latter is semi-diurnal having a smaller 
tidal range of 0.2-1.0 m. 


MIXED nor 





Fig 2.7: Distribution of tidal types in Southeast Asia (after Wyrtki, 1961) 

The level of water in any particular coastal system is the 
combined result of tidal forces that may be modified due to the nature 
and extent of the estuaries, coastal configuration, and terrigenous 
river discharge. The response to these forces can be determined from 
an analysis of tide gauge records. 

Tidal amplitude varies from place to place. A marine tidal chart 
will give a good idea of tidal conditions. Where the amplitude is 
high, the area subject to periodic tidal flushing is corresponding 
large and usually gives rise to a wide range of ecological sites. This 
is particularly so in areas where the continental shelf slopes gently 
towards the sea and the coastal current is weak enough to allow the 
mangrove formation to extend seawards. Generally in the Caribbean 
area, due to the lower tidal amplitude, the range of ecological sites 
is restricted, and accordingly, the vegetational types are relatively 
smaller in numbers and extent. 

2.5.2 Salinity 

A saline environment is required for stable mangrove ecosystems, 
as many species are less competitive under non-saline conditions (Lugo, 
A.E., 1980) . 


However hypcrsalinity can adversely affect mangroves and a given 
site is considered to be hyper saline when the salinity (surface or 
interstitial soil levels; exceeds that prevailing in the sea (In most 
areas this level averages 35 ppt) . The effect due to salinity and the 
resulting strongly negative osmotic pressure of soil water, is a 
progressive stunting of the mangrove canopy inland from the water's 
edge. This can be almost universally recognized and takes place 
regardless of species composition. 

Zonation of species is partly influenced by salinity, although 
the extent of its influence depends on local climatic and edaphic 
factors (Chapman, V.J., loc.cit.; Mukherjee B.B. and Mukherjee, J., 
1978; Watson, J.G., 1928; Hann, J.H. de, 1931; Frodin, D., 1985; Thorn, 
B.Q. 1967). Similarly, Saenger et al. (1983) conclude that the 
efficiency with which each species deals with high soil salinities 
largely determines its position in the intertidal zone. 

Mangroves are considered to be facultative halpphytes, i.e. they 
can often survive though not necessarily thrive in non- saline habitats 
(Cintron, G. and Schaeffer-Novelli, Y., 1983a: Walsh, G.B., 1974). It 
has often been reported that the growth of many halophytes is depressed 
without sodium chloride in the external environment (Jennins, 1970; 
Flowers et al., 1977; Greenway and Munns, 1980). In the case of 
mangroves a number of studies now available point to a similar 
physiocheraical trend and support the hypothesis that the presence of 
limited amounts of sodium chloride in the external medium is required. 
Table 2.5 summarises the data for 2 species. 

Table 2.5: Effect of salinity on maximum growth of mangroves 


Percent of i 
seawater (%) 


Avicennia marina 
Avicennia marina 

Avicennia marina 
Avicennia marina 
Avicennia marina 

Rhizophora mangle 
Rhizophora mangle 
Rhizophora mangle 


10 - 50% 



Connor, 1969 
Clarke and Hannon, 

Downtown, 1982 
Clough, 1984 
Burchett, Field and 
Pulkownik, 1984 
Stern and Voigt, 1959 
Pannier, 1959 
Clough, 1984 

Note: i/ Concentration of NaCl required for maximum growth / 

/ The growth criterion uaed, i* the accumulation of total dry weight. 

Despite the apparent differences in salinity levels recorded 
above, which may be due to seasonal variance of the seawater salinity, 
there is an optimal salinity range for maximum growth. At extreme 
levels mangrove species suffer damage and even mortality. The die-back 
of Sundri (Heri tiers femes) has been ascribed to an adverse increase in 
soil salinity (Christensen, B. and Snedaker, S.C., 1984; Chaffey, D.R. 
et al., 1985) . Different range of salinity values are, however, quoted 
by different authors. (Cintron, O.Y., and Schaeffer-Novelli, Y., 
1983b; So to, R. and Jimenez, J.A., 1982). 


In afforestation where matching species to site is required, the 
salinity range required for optimum plant growth and regeneration for 
any given species is more useful than its wider 'ecological' survival 
limits. Yeo and Flowers (1980) have stressed that the phenomenon of 
growth response to an increase in salinity should be considered quite 
separately from the tolerance to extreme salinities, which must be 
considered to be much higher than the optimal salinity for growth. 
Information on ecological tolerance limits are nevertheless important 
in predicting the continuance or displacement of species growing under 
altered or high-stress habitats. In this respect, when rehabilitation 
programmes are carried out in ecologically degraded areas that may be 
hypersaline, it is prudent to use propagules collected from individuals 
that grow in similar sites, because there may be ecotypes or hybrids 
that are better adapted to such stressed environments. Jimenez and 
Soto, (1985) have prepared a list of Pacific coast Central American 
mangrove species and their maxima salinities, including herbaceous 
plants such as Hymenocalis littoralis and H. pedalis (Liliaceac) . 
HymenocaliB sp. appear to be good site indicators for habitats that are 
relatively well flushed and which have a low salinity. These sites are 
optimum for Rhizophora mangle and Pelliceria rhizophorae. 

Salinity is also an important physiological factor affecting 
marine and brackish water animals, and is used as one of the parameters 
to assess aquaculture potential. Very high or low salinity will affect 
growth, and if extreme can be lethal. Salinity of less than lOfc can 
produce an off-flavour in shrimps (Clifford, H.C., 1985). Penaeus 
vannamei, one of the most desirable species for culture, does best at 
salinities ranging from 15 to 20fc. According to Midget (1985) the 
salinity-growth relationships for most penaeid shrimp species are as 

Table 2.6: Salinity-growth relationship for penaeid shrimp 


5-15 4 30-40 

2.5.3 Other ednohic factors 

Mangrove soils are usually alluvial. They are normally 
featureless, hydromorphic with varying degree of gleying in the 
subsoil horizons. Troll and Dragendorff (1931) considered that the 
black colour of many mangrove muds is produced by anaerobic bacteria 
reducing sulphates to sulphides. 

The acid sulphate problem 

The high organic and iron content in mangrove soils combined with 
the ever present sulphate from tidal seawater renders them particularly 
susceptible to acid sulpha t ion due to oxidization, as often happens 
during pond construction. See Box 2.2 on the following page. 


A^jif A ^aj^^laM -** d^i MMI ^w>^>^ _.^..w**.,.i..,.-,^- .i^;^.^, 

*nQ MIHUUII CT InOM pyfRl lloUtiiniB IMO 

kttortton wd to tormrton of Sutfurte add 880, whtoh 

Box 2.2: The add sulphate problem 

When this occurs, the pond pH often falls to 3 or less - a 
condition that presents problems for both aquaculture and agriculture 
(Watts, J.C.D., 1969, and Kanapathy, K., 1975). Potter (1977) cites 
low natural production, poor response to fertilizers, and slow fish 
growth as some of the effects of using such soils for fish culture. 
Cook and Rabanal (1978) recommend liming for soil pH below 6.5 but the 
cost may be prohibitive, as about 30 t/ha of limestone may be required 
to increase the upper 30 cm of an acidic mangrove soil by only one pH 
unit (Kanapathy, K., loc.cit.). The potential danger of acid 
sulphation should be considered in the physical conversion of mangrove 
soils, as well as the threat of acid contamination to the environment 
and fishery. Dunn (1965) reported the mass killing of fish in some 
cases where heavy rains washed the soil acids into rivers. 

The problem of acid sulphate soils in relation to conversion of 
mangrove soils to salt ponds and rice fields has also been studied by 
Thomlinson, (1957) and Hesse (196 la and 1961b) . In Sierra Leone, where 
rice has been cultivated in mangrove areas since 1855, Hesse found that 
soils previously covered with Rhizophora sp. tend to develop soil 
conditions adverse to rice growth due to these soils being sulphidic 
and highly fibrous in nature, which results in the formation of acidic 
sulphate and the release of aluminium ions when empoldered. Soils 
previously covered by Avicennia on the other hand, were nonr fibrous and 
did not pose any problems when empoldered. One of the reasons behind 
these findings may be that in Sierra Leone, Avicennia is usually found 
on sandy soils, which generally have a lower content of phosphorus and 
oxidisable sulphur than clayey soils, often dominated by Rhizophora sp. 

Soil sampling provides an idea of the nature of the soils which 
are texturally influenced by silt deposits within the estuary. For 
example the high silt transport of the Terraba river precludes or 
reduces the significant development of benthic communities in the 
rivers, which may have an effect on mangroves via the reduced rate of 
turnover of the soil, and on aquatic productivity which depends on 
phytoplankton production and the organic detritus from the mangrove 
forest . 

Soil characteristics affect natural and artificial regeneration. 
Plant species growing in the intertidal zone are exposed to many 
stresses . 


the high salinity (and its fluctuations) creates 
physiological stress. Secondly, water-logged soils have a low content 
of interstitial oxygen. Such anaerobic conditions compel the plants to 
get oxygen either from the air or from the very top layer of the soil. 
In the less aerated soils, the Rhizophora spp take on a different 
physiognomic form with large numbers of hanging aerial roots that 
originate from both the stem and upper branches, as well as, lateral 
"running" roots. Aerial roots serve not only an aeration function but 
also stabilise the stems as "prop" roots, although these functions vary 
according to species (Percival, M,, and Womersley, J.S. 1975). 
Thirdly, fluidity of the soil substrate is still another major 
constraint, especially for tree species. 


2.6.1 Vegetal formations and communities 

In Central America, Jimenez and Soto (1985) recognized 3 mangrove 
zones along the Pacific Coast in Costa Rica, viz. North, Central and 
South Pacific. The vegetation is grouped into 3 types according to 
their distribution, biological characteristics, soil salinity and 
inundation intensity as follows: 

(i) Nuclear vegetation 

This type constitutes the mangrove forests sensu stricto as it 
comprises species in the intertidal zone that are dependent on saline 
influences, the so-called obligate halophytes. Most species have 
special adaptations which enable them to grow in the mangrove 
substrate, such as vivipary, high salt tolerance, ability to withstand 
tidal submersion, pneumatophore or aerating roots, succulence and salt 
excreting glands. 

The 5 most important species are Rhizophora mangle L., Rbizophora 
harrisonii Leechman (Rhizophoraceae) , Pelliciera rhizophorae Triana and 
Planchon (Pelliceriaceae) , Avicennia germinans L. (Avicenniaceae) and 
Laguncularia racemosa L. Gaertn. (Combretaceae) . 

(ii) Marginal vegetation 

The species here are commonly associated with the mangroves in 
the landward fringe, in seasonal freshwater swamps, beaches and/or 
marginal mangrove habitats. Though they exist in the mangroves, these 
species are not restricted to the littoral zone. Conocaipus erecta 
(Combretaceae) is not found in the mangrove proper. Jfora oleifera 
(Triana) Duke (Leguminosae) is abundant in the south Pacific coast, 
particularly in Peninsula de Osa, where it grows in seasonal swamps 
that may be quite saline (25fc) . Other species present include Annona 
glabra L. (Annonaceae) , Pterocaxpus officinaiis Jacq. (Leguminosae), 
Hibiscus tiliaceus L. and Pavonia spicata Killip (Malvaceae) . 

The fern Acrostichum aureum L. (Polipodiaceae) negraforra is very 
extensive in the brackish water zone and poses a threat to seedling 


(iii) Marginal facultative vegetation 

Carapa guianensis (Meliaceae) grows in the South and partly South of 
the central Pacific coast in salinity around lOfc. Other species are 
Elaeis oleifera and Raphia taedigera. This is the inter- terrestrial 
zone which in the more equatorial climate would correspond to the 
Melaleuca leucadendron swamps (e.g. Southern Vietnam) . This 
vegetational type has limited forestry potential. It is highly 
modified due to human development and more suited for other land uses. 

Lugo and Snedaker (1974) identified and classified mangroves 
according to six community types based on forest appearance, and 
related to geological and hydrological processes. Each type has it own 
characteristic set of environmental variables such as soil type and 
depth, soil salinity range/ and flushing rates. Each community range 
has characteristic ranges of primary production, litter decomposition 
and carbon export along with differences in nutrient recycling rates, 
and community components. A brief description of the community types 
based on the Floridian experience as shown in Figure 2.8 is as follows: 

( 1 ) Overwash mangrove forests - the red mangrove is the dominant species on these 
islands that are frequently inundated and flushed by the tides, resulting in high rates of organic 
export. Maximum height of trees is about 7 m (23 ft). 

(2) Fringe mangrove forests 
these mangrove fringes are found along 
waterways, best defined along shorelines whose 
elevations are higher than mean high tide levels. 
Maximum hight of mangroves is about 10m (32 

(3) Riverine mangrove forests - 
this type may be tall forests along tidal rivers 
and creeks, subject to regular flushing. All the 
three Floridian mangroves, viz., White 
(Laguncularia racemosa), black (Avfcennia 
germ/nans) and red mangroves (Rhizophora 
mangle] are present. Stand height may reach 1 8 
- 20 m (60 - 65 ft). 

( 4 ) Basin mangrove forests - this 
generally stunted type is located in the interior of 
swamps in depressions channelling terrestrial 
runoff toward the coast. Red mangroves are 
present where there is tidal flushing but towards 
the inland portion white and black mangroves 
predominate. Trees may reach 15m (49 ft) in 

(5) Hammock forests - generally 
similar to type (4) above but they are found on 
slightly elevated sites relative to surrounding 
areas. AH species are present but the height is 
seldom more than 5 m (16 ft). 








fig 2.8: Mangrove community types 

Lugo & Snedaker, 1974 


(6 ) Scrub or dwarf forests - this community type is typicaHy found in the flat coastal 
fringe of south Florida and Florida Keys. All three species are found but rarely exceed 1.6 m (4.9 ft). 
Nutrient appears to be the limiting factor. 

In Vietnam, the better mangroves 'are confined to the relatively 
sheltered coasts in the South, with the best mangroves located in the 
Ca Mau peninsula. San and Hong (1984) classified these mangroves into 
4 principal distribution zones. The principal pioneer species are 
Avicennia lanata, Sonneratia caseolaris, S. alba and Avicennia alba. 

Vegetational changes away from the waterways and associated 
values of water properties in Thailand have been studied by Aksornkoae 
(1975) and are shown in Figure 2.9. 

2.6.2 Zonation and inundation 

While air and water temperature determine the latitudinal limits 
of mangrove species, rainfall generally governs the distribution and 
zonation of mangrove species along many non-mountainous coasts (Blasco, 
F. 1984) . In Australia, temperature and water balance are considered 
more important. 

Macnae (1966) attributed the distribution of mangrove trees and 
hence their zonation to the interaction of, (a) frequency of tidal 
floodings, (b) salinity of soil water; and, (c) water logging of the 
soil (drainage) . Walter and Steiner (1936) consider the degree of 
flooding, soil nature and salinity as important factors. With respect 
to tides, Chapman (1976) considers that the most important factor is 
the number of consecutive days with no tidal flushing. 

Whereas the degree of flooding, which depends on soil level, is 
important in the establishment and dispersal of propagules, its effect 
on mature stands may be less pronounced. Rabinowitz (1978) has 
suggested that the morphology of the propagules controls the zonation 
of mangroves in Panama because smaller propagules can be transported 
further inland through already established vegetation by tides. 

Although there are many descriptions of the contributory factors 
that account for zonation (cf. Floyd, 1977; Paijmans and Rollet 1977), 
it is Lugo (1980), Woodroffe (1983) and, in Papua New Guinea, Johnstone 
and Frodin (1982) who have attempted more thorough analyses. Johnstone 
and Frodin have proposed six types of likely causes: 

Inundation and depth of water 

Wave action 


Salinity/freshwater regime 


Biota and biotic interactions 

Some or all of the above factors have been emphasised by 
different authors, but the last-named factor has often been neglected. 


CM *- CM 

CM f- - 


OJ p- - 


o> o 

in P% r 

oo in do 

cvj ^ 


ev o 

C* P- 


CM CNJ r- 

S 55 .s, 



ocMCOr-tfih*<* in 

CM r- 

rv CD 
__>io co co 

^< O) ^ O% CD C9^ CO ^^ 

GO U>r-^CM COO CM *t 

CM CM r cn 

^^S 8 

b or*CMor*f- 

Li. E CM CM r- 




O* in 
^ror- o in 

_ _ . MOP- CM f^ 

CM CM P 00 

^ ^* ^^ WM 

I O O 

^ T 

+* >i 

41 C O 

X ^_ ui * 

B "~ 42 8 o*" o 






Vegetational patterns or zones are often easily recognisable in 
mangroves. Zonation studies provide a useful indication of the 
ecological and silvicultural management requirements of the forest 
stands, particularly in selecting the appropriate habitats for 
preferred species and in evaluating in situ stand productivity under 
natural conditions. Regeneration problems can often be reduced by 
avoiding the promotion of species outside their natural habitats. In 
peninsular Malaysia, for instance, where intensive mangrove management 
has been practised since 1902, the refinement of the floristic 
composition, with emphasis on a few preferred species, has spawn many 
regeneration problems. Rhizophora apiculata is planted in landward 
zones that are marginal for its development, and, as a result, natural 
regeneration is often inadequate after clear- felling, and an extensive 
afforestation programme is required. 

Three classifications, inter alia, are usually used to describe 
zonation in mangroves. Two of these are based on the gradient concept 
while the third follows the Braun-Blanquet system broken down into 
orders, alliances and associations (Chapman, 1976) . 

(i) Watson (1928) in a pioneering study on the ecology of 
Malaysian mangroves, divided the West Malaysian mangrove communities 
into five classes based on the frequency of inundation. The 
silvicultural significance of this classification is that a species is 
allotted to a particular inundation class based on its ability to 
regenerate itself, and not merely being present. This type of work has 
been extended to include an appreciation of the functional properties 
of different species and assemblages of species along environmental 
gradients (Lugo and Snedaker, 1974) . 

(ii) De Hann (1931) considered salinity as the primary factor in 
controlling distribution and tidal inundation as a subsidiary factor. 
His scheme has two main divisions each with subdivisions as follows: 

A. A brackish to saltwater zone with salinities at flood tide 
of between 10 - 3Q& and Inundated 

Al once or twice dally on each of the 20 days/month 

A2 10-19 times per month 

A3 9 times or less per month 

A4 only few days per month 

B. A fresh to brackish water zone with salinities between - 1(K 

Bl more or less under tidal influence 
B2 seasonally flooded. 

The relationship between Watson's and De Haan's classifications 
is compared in Tabl* 2.7 below. 

Determining tidal inundation 

Three methods may be used to determine the position of forest 
communities and species in relation to tidal levels : 

(a) conduct a local levelling survey based on a bench-mark above 
high- tide level, this bench mark being referenced to a tide level 
in the nearest harbour, 

Table 2.7t b 


s and De Hum's clMriflcatton 

Claee flooded by 


Inundation frequency 



All high tides 
Medium high tide. 
Normal high tidee 
Spring high tidea 
Abnormal or 
equinoctial tidea 
Wet aeaaon only 



11 - 13 

13 - 15 

(b) a levelling survey based on a tidal pole established in the swamp 
and related to the nearest harbour tidal gauge, or 


establish a local tide gauge. 

However, a simpler method, which may prove sufficient for most 
situations, is to tie a number of horizontally tilted bottles one on 
top of another in a vertical array held in place by two small stakes 
and record the tidal levels by noting the height of the topmost bottle 
filled with sea-water. Care should be taken to ensure that the 
measuring device is located away from exposed positions. Once the 
spring tide height is determined, a permanent tidal pole may be 
established. Again by means of local level surveys it is then possible 
to extrapolate this crude bench mark to other positions. 

A knowledge of the depth and frequency of flooding is useful in 
plantation establishment, where special attention should be paid to 
areas with inadequate tidal flushing or subject to deep flooding. The 
above method can also be used to estimate the relative amounts of silt 
deposition between sites. For example if several of the recording 
devices are constructed and placed in different sites, water turbidity 
can easily be compared. It can also be used to gauge the effects of 
logging operations on substrate stability, by observing the quality of 
the water collected in the bottles before, during and after logging 

The position of some New and Old World species with respect to 
Watson's and De Haan's classifications is summarized in Table) 2.8 on 
the following page. 

(iii) Walter and Steiner (1936) working on mangroves at Tanga in 
Tanzania named the zones after the dominant trees. This scheme was 
adopted and modified by Macnae (1966) , who recognized the following 
zones : 

(a) Landward fringe 

(b) Zone of Ceriops thickets 

(c) Zone of Brugruiera forests 

(d) Zone of Rbizophora forests 

(e) Seaward Avicennia zone 
(d) Sonneratia zone 


These zonation patterns are not universal and are modified by 
climate, salinity, coastal morphology and freshwater outflow. The 
Sundarbans forest in the deltaic plains of the Ganges and Brahmaputra 
rivers falls within the "landward fringe" category, the most variable 

Table 2.8 New and Old World species 

according to 

(after Chapman, 197S) 

Watson's inundation 
classes (1928)- 
Type of flooding 

de Haan's inundation 
classes (1931) based 
on salinity/frequency 

Old World dominant 


New World 

A. Brackish to saline 

1-3% salinity at 

high tide 

1 . All high tides 

A1. 1-2 times/day; at 

Sonnerat/a alba 



least 20 days/month 

S apetala 

700 + 


2. Medium high tide 

A2. 10- 19 days/month 





per year 


3. Normal high tides 

A3. 9 days/month 

landward fringe 



4. Spring tides only 

A4. Only a few 






per year 




5. Storm high tides only 

B. Fresh to brackish 


4- 100 

SaJ/na or 

water; salinity 0-1 0%o 

or salt flats 

per year 



B1 . More or less under 

Nypa fruticans 

tidal influence 

. .>4 J 

The use of inundation classes in this report does not imply a 
successional relationship but rather highlights the usefulness of such 
a classification in forest management, where a knowledge of tidal 
conditions and depth of flooding is required in ^Jhe planning for 
harvesting methods and reforestation. In Costa Rica, for example, the 
frequent occurrence of R. mangle in wet and softer soils precludes the 
use of tramways for wood extraction, whereas A. germinans stands in the 
drier and more consolidated soils can often be accessed from the 
landward side using semi -mechanized harvesting systems as the soil has 
a higher load-bearing capacity. Artificial regeneration in Watson's 
Inundation Class I (W.I.C-I) is usually difficult because of the depth 
of the inundation, its frequency, and very soft immature soils. 

2.6.3 Succession - ecological aspects 

Succession refers to the replacement of the biota of an area by 
one of a different nature. There has been much debate in the 
literature concerning the relationship between zonation and succession 
(Lugo, 1980) . 


The idea that {zonation recapitulates succession (Davis, 1940), 
implying an inevitable development towards a non-mangrove 'terrestrial' 
vegetation type has been much criticised (e.g. Rabinowitz, 1978) as 
such a progression of events is not universal. (Johnstone and Frodin, 
1982; Lugo and Snedaker, 1974) . 

However, the relationship between zonation and succession has 
been clearly documented in mangroves with prograding shorelines, high 
rainfall during all seasons and considerable intertidal range (Watson, 
1928; Putz and' Chan, 1986). 

The formation of newly accreted mud- flats in prograding 
shorelines gives rise to a permanent ecological transfer of sites on 
the landward fringe to dryland forest above tidal influences. However, 
this successional model emphasizes the physiographic role of mud in the 
formation of accrescent shorelines rather than the classical model of 
mangroves as land-builders. 

Geomorphic processes and severe, episodic climate events 
determine the status of the intertidal habitats. Thorn (1984) has 
suggested that a system for classifying mangrove shores into 
prograding, eroding or stable states is of fundamental importance in 
understanding present vegetational distributional patterns within a 
local region, as well as the time scale of vegetation dynamics. The 
prograding shoreline is only one of many geomorphic landforms, 
therefore, the successional model cannot be universally applied to all 
coastal ecosystems. 

Putz and Chan (1986) in an analysis of stand growth and dynamics 
in a mature forest in Malaysia, monitored since 1920, conclude that 
observed species replacement patterns resemble the classical 
successional process observed and propounded by Watson. Rhizophora 
apiculata, a fast growing but less shade tolerant species is gradually 
being replaced by another fast growing but shade tolerant species, 
viz., Brugruiera gymnorhiza (Rhizophoraceae) as well as, other typically 
more landward species: The same investigators suggest that shade 
tolerance and dispersal characteristics should be included among the 
ecological factors influencing the distribution of tree species in 
mangrove forests. 

The Terraba-Sierpe mangroves along the Pacific coast of Costa 
Rica and the Ayeyarwady mangroves in Myanmar are example of wave- 
dominated delta systems (sensu Thorn, 1984) . The high upland silt load 
would normally produce an accrescent shoreline had it not been 
physically checked by the erosive actions of waves and tidal currents 
at the river mouths. This leads to the formation of elongated shoals 
and sand bars and rounded shorelines. The high silt load of the 
terrigenous run-off provides an annual silting of the forests, 
contributing to soil fertility and forest productivity. At Playa Garza 
in Costa Rica, mechanical analyses of soil samples indicate that 
samples located close to the river mouth have a higher percent of sand, 
which usually improves soil firmness and drainage (aeration) . Such 
sites carry good stands of R harrisonii and P. rhizophorae. 


Where the tidal environment is very fluid and dynamic, the rate 
of physical environmental change in ; localised sites may approximate or 
even exceed endogenous ecological processes of change, (cyclical 
changes, gap regeneration or succession). In this respect mangrove 
vegetation dynamics must be viewed in A broader perspective than that 
accorded to terrestrial vegetation where environments are normally 
stable in relation to such processes. 

Many ecological studies on vegetational pattern rely on species- 
level analysis (floristic attributes alone) to show segregation and 
development of dominance, based on the implicit assumption that 
different species have sufficiently different niche requirements (sensu 
Grubb, 1977) . There is little doubt that species shifting takes place. 

In the field, this can be observed by an initial change in 
individual life form collectively expressed as changing community 
structure. Silviculturally, some of the significant changes are i) 
potential above-ground biomass (site quality/productivity), ii) tree 
species composition (site specificity) and iii) community structure 
(form quality) . 

Although there appears to be an optimum range of sites for the 
principal species, and interspecific competition may be low due to the 
small number of species, their physiognomic form may vary from high 
forest to thickets or scrubs. The probable presence of interbedded 
lenses of alluvium is another factor that causes changes in micrp- 
habitat and there are many processes which are not well understood, 
such as for example that of 'deep flooding' in some areas (Noakes, 
D.S.P. 1956 in p. 187) and the occurrence of 'estructuras circulares de 
vegetation' in Gabon (Legigre, J.M. 1983 in p. 20) . The latter may be 
caused by lightning and/or incidence of commencing subsidence. In the 
Matang area along the west coast of Peninsular Malaysia, circular 
blanks due to trees killed by lightning are quite common especially 
towards the seaward side. 

In Malaysia the conditions under which the Rhizophora app form 
the bulk of the economic growing stock may be regarded as the optimum 
stage of development of the mangrove swamp. These conditions are 
briefly i) inundation by ordinary tides, ii) moderate salinity (20- 
35fc) , iii) aeration and enrichment of the soil through organic 
accumulation and iv) an abundance of small streams and gullies to 
assist propagule dispersal and promote efficient tidal flushing. The 
natural regeneration potential is usually good. The mangroves and 
their surrounding mudflats or lagoons are ideal sanctuaries for various 
crustacean species. There is usually also an active crab population, 
principally the fiddler crab, t/ca sp and various species of Sesarma. 
Their innumerable burrows help to aerate the subsoil. Some organic 
matter is also transferred into the subsurface soil, consequently such 
soils may have a higher level of organic matter. It is interesting to 
note that the better forests usually do not have festoons of hanging 
aerial roots, and this is in agreement with Watson's observation that 
stilt root development is more accentuated in inferior soils or deep 
flooded area. 


In Central America, R. mangle may be regarded as the principal 
pioneer species in the mangroves sensu s trie to. As it is relatively 
intolerant of shade it is normally found in the immature and softer 
soils along sunny river banks, where it regenerates freely. Where the 
soil is firm and more elevated, it is replaced by other species such as 
Pelliciera rhizophorae, which is more shade tolerant and can be found 
on quite sandy soils. In the less sunny and firmer creeks and 
channels, the gallery forest may comprise a mixture of R. mangle and R. 
harrisonii. In newly formed mud- flats Laguncularia racemoaa sometimes 
is the pioneering species. The landward fringe is a very variable zone 
and does not contain many mangrove species of economic importance. The 
principal species is Avicennia germinans, which generally has poor 
form. L. racemoaa and P. rhizophorae also occur in this area. 

In SB Asia, the pioneering species are usually Avicennia lanata 
and alba or Sonneratia alba. 

2.7.1 Wildlife 

A compilation of wildlife species (mammals, reptiles and 
amphibians) and birds found in the Sundarbans, the world's greatest 
contiguous mangrove area, is presented in Das and Sidddiqi (1985) . 
McNae (1968) gives a general account of the fauna of mangroves in the 
Indo West Pacific Region and Saenger et al. (1983) look at the global 
status of mangrove ecosystems including their fauna. 


Many mammals frequent mangrove 
habitats but only a few live there 
permanently and fewer are restricted 
to them (FAQ, 1982) . In many 
countries however, the mangroves 
represent the last refuge for a number 
of rare and endagered mammals. 

During low tide monkeys (Macacus 
irus) are commonly seen foraging for 
shell -fish and crabs in Malaysia, and 
the white* faced monkey (CeJbus 
capucinus) feed on pianguas (cockles) 
in Costa Rican mangroves. Reportedly, 
where such monkeys are numerous, the 
area is poor in pianguas. They also 
do a certain amount of damage to newly 
established seedlings by uprooting 
them. The Malaysian proboscis monkey 
(Nasal is larvatus) is endemic to the 
mangroves on Borneo, where it feeds on 
the foliage of Sonneratia caseolaris 
andlttpa frutican* (FAD, 1982) as well Fig. 2.10: 
as on Khizqphora propagules. The 
monkeys, in return, are preyed on by 
the crocodiles and hunted by poachers. 

RMzophora propafde. 


Other mammals include the Royal Bengal Tiger (Panthera tigrris) , 
the leopard (Panthera pardus) and the spotted deer (Axis axis) in the 
Sundarbans; wild pigs (Sue scrofa) and mousedeer (Tr&gulus ap) in Nipa 
swamps throughout South and Southeast Asia; and small carnivores such 
as fishing cats (Felix viverrima) , civets (Viverra sp and Vivererricula 
sp) and mongooses ( Herpes tes sp) . Otters (Aonyx cinera and Lutra sp) 
are common, but rarely seen. 

Dolphins, such as the Oangetic dolphin (Platanista grangetica) and 
the common dolphin (Delphinus delphis) are also found in the rivers of 
mangroves, as are Manatees (Trichechus senegralensis and Trichechus 
manatus latirostris) and Dugongs (Dugong dugon) although these species 
are becomming increasingly rare and in many places are threatened with 

Reptiles and amphibians 

Crocodiles and alligators are some of the most significant 
reptiles that naturally inhabit marine and estuarine environments. 

Two species Crocodilus a cut us (lagarto) , Caiman crocodilus 
(largarto cuajipal) are found in Costa Rica, where they are listed as 
endangered species, largely due to international trade in their hides. 
C. acutus has a very wide geographic range and is found in Cuba, 
Pacific Coasts of Central America, Florida and Venezuela. The Cuban 
species, Crocodilus rhombifer is found in Cienaga de Lanier and is 
endemic. The American alligator Alligator mississippiensis is listed 
as endangered in Florida (Hamilton and Snedaker, 1984) . In West Africa 
the Long Snouted Crocodile (Crocodilus cataphractue) is found in 
mangrove areas and in Asia the saltwater crocodile Crocodilus porosus 
is endangered over a large part of its range. Efforts are, however, 
being made to conserve it in India, Bangladesh, Papua New Guinea and 
Australia (FAO, 1982) . 

The large lizards, Iguana igruana (iguana) and Cetenosaura similis 
(gar robe) are commonly found in the mangroves in Latin America, where 
they are eaten by the local people, as are their cousins in West Africa 
(Varanus exanthema ticus) and Asia (Varanus sal va tor) . 

Riverine tortoises are common and marine turtles are known to lay 
their eggs on the sandy beaches in many mangrove areas throughout the 
world. Along the Pacific coast of Costa Rica, the two most important 
egg-laying sites visited by the Pacific Ridley Turtle (Lepidochelys 
olivacea) is in the Playa Nacite in the Santa Rosa National Park and in 
the Playa de Ostional near the Rio Nosara. This turtle is relished for 
its meat and weighs an average of 40 kg. Its numbers are diminishing 
because of predation and over- exploitation in some countries, notably 
Mexico and Bguador. 

A number of snakes can also be found in mangrove areas especially 
in the landward fringe. 


Figure 2.1 1: The spotted deer (Axis axis) in the Sundarbans. 

Photo by M.L.Wilkie 

figure 2.12: The saltwater crocodile (Crocodilus porosus) in Indonesia. 

Photo by M.L.Wilkie 


2.7.2 Avifauna 

The tidal swamp is an ideal sanctuary for avifauna, some of which 
are migratory. According to Saenger et al. (1984), the total list of 
mangrove bird species in each of the main biogeographical regions 
include from 150 to 250 species. Worlwide, 65 of these are listed as 
endangered or vulnerable. Several surveys of avifauna in mangrove 
areas in South East Asia have been carried out. (See for instance Das 
and Siddiqi 1985; Erftemeijer, Balen and Djuharsa, 1988; Howes, 1986 
and Silvius, Chan and Shamsudin, 1987.) 

In Cuba, there are several endemic species which occupy highly 
specialized ecological niches such as the canario del manglar 
(Dendroica petechis gundlachi) and the smaller oca del manglar (Rallus 
longirostris cariJbaeus) . 

The most numerous birds are 
the waders, herons, egrets and 

Birds of prey include the 
sea eagles (Haliaetus 
leucogaster) , brahminy kites 
(Haliastur Indus) , ospreys 
(Pandiorz haliaetus) and fish 
eagles (Ichthyphagus ichthyaetus) . 

Kingfishers and bee-eaters 
are among the most colourful birds 
commonly observed in mangroves. 

2.7.3 Aquatic resources 

Fig. 2.13: Large Sandplover - by D. Beadle. 

The importance of the mangrove areas as feeding, breeding and 
nursery grounds for numerous commercial fish and shellfish is well 
established (Heald and Odum, 1970; MacNae, 1974; Martosubroto and 
Naamin, 1977) . Similarly, Chong (1987) reported that the location of 
fishing grounds in Sierra Leone is geographically correlated to the 
distribution of coastal mangroves. See also Tabl* 3.7 on page 63. 

The development of soft clayish mud, where the crabs can make 
their burrows, and the growth of seagrass or turtle grass have been 
observed to attract the crustacean fauna along the mangrove areas. 

Matthes and Kapetsky (1988) have prepared a worldwide compendium 
of mangrove-associated aquatic species of economic importance including 
such information as geographical extension of each species; the parts 
of the mangroves in which it is found; the organism's dependency upon 
the mangroves and its quality /use in fisheries. 


There are reported to be over 120 species of fish caught by 
fishermen in the Sundarbans (Seidensticker and Hai, 1983), almost all 
of which are brackish water and esuarine species. 


Ulloa (1978) reported 92 species of fish belonging to 13 families 
caught in Jiquilisco Bay in El Salvador. 

Species of commercial interest include mullets (Nugilidae) , 
snappers (Lutjanidae) , milkf ish (Chanos chanos) , sea bass (Lates 
calcarifer) and tilapia (Cichlidae) . The most conspicuous fish is 
perhaps the muds kipper (Periopbthalmus sp.) f which is endemic to the 
mangroves . 


Despite the presence of the more spectacular mammals and 
reptiles, indications are that the animals which contribute the 
greatest biomass in the mangroves are the shellfish [a collective term 
for crustaceans (crabs and prawns) and molluscs (bivalves and 
gastropods) ] . 

The fiddler crab, Uca sp and the various species of Sesarma are 
common inhabitants in the intertidal mangrove zones throughout the 
Indo- Pacific region. Crabs of the Portunidae family have been observed 
in the Arabian Gulf area and the United Emirates. 

The edible crabs (Scylla serrata in Asia and East Africa and 
Callinectes latimanus in West Africa) are a highly valued mangrove 
product . 

The most common prawns include the giant freshwater prawn 
(Macrobrachium rosenbergii) and the marine penaeid prawns (Penaeus 
indicus, P. merguiensis , P. monodon, Metapenaeus brevicornis) . All of 
these species probably have a similar basic life history with spawning 
occurring offshore, an inshore migration of larvae, an estuarine 
juvenile stage followed by an offshore breeding migration to complete 
their biological life-cycle. However, the species differ in the extent 
to which they move offshore during this migration. Surveys in Malaysia 
showed that the genus Penaeus was abundant across all depths up to 50m, 
while Metapenaeus .were most abundant in the 11-30 m range and 
Parapenaeopsis were more restricted to the 5-20 m zone. 

Reportedly, penaeid shrimps off the coast breed throughout most 
of the year but with observed peak-periods during May-July and October- 
December, which coincide with the coming of the monsoons. In Western 
Malaysia peak ingress of . me ran lens is postlarvae was reported during 
November and December. 

After three to four months in mangrove estuaries, juvenile 
shrimps migrate into the shallow coastal waters from March to June 
where sexual maturity takes place. When larger, they move further 
offshore to spawning grounds in depths exceeding 10 fathoms. Major 
spawning migrations begin in June and continue to late January. 

Regarding molluscs, in Central America, the larger bivalve 
Anadara grandis (chucheca) , is now rare due to over- exploitation. The 
smaller ark clams 'pianguas' comprising principally two species Anadara 
multicostata and A. tuberculosa are now exploited in place of chucheca. 
Anadara tuberculosa, Sowerby, is the molluscan bivalve commonly found 
in mangrove ecosystem from Lower California to Peru (Keen, 1971) . 


The most important bivalve in the I ndo- Malayan mangroves is the 
blood cockle (Anadara grranosa) and gastropods commonly collected 
include Cerithidia obtusa, Telescopiwn mauritsii and T. telescopium. 

Oysters are also important sources of aquatic production, which 
like shellfish can be cultured provided suitable substratum is provided 
to attract the spats and the estuarine conditions are right. 

The importance of shellfish as a source of readily accessible 
protein and an economic renewable resource for coastal dwellers makes 
it the single most important exploited species in the mangroves. 


Until recently, studies on the benthic fauna of inter-tidal 
mudflats in tropical regions have been exceptionally rare in spite of 
the fact that such studies would provide indications of the value of 
mudflats as potential feeding habitats for water birds and marine and 
estuarine fish (Erftemeijer, Balen and Djuharsa, 1988; Silvius, Chan 
and Shamsudin, 1987) . Recent studies have however been carried out for 
instance in Hong Kong, China and Taiwan - often in connection with 
feasibility studies on the use of mangroves for waste water treatment 
(refer to HKUST, 1993) . 

Benthic fauna includes juvenile fish, crustaceans, crabs and 
bivalves and are divided into two classes : Macro-benthos > l mm and 
Meio-benthos < 1 mm. 


The successful integrated management of mangrove wood and non-wood 
resources depends on an understanding of, firstly, the ecological and 
silvicultural parameters for forest management (primary production) and 
secondly, the biological role that the primary production from the forests 
plays in the mangrove food web of aquatic resources (secondary production) . 
An understanding of the role of key species in maintaining the equlibrium of 
a particular ecosystem is likewise essential. 

2.8.1 The food web 

Our present knowledge on energy flow in mangrove ecosystem is 
mainly based on the pioneering work on food chains in Florida (Heald, 
1971; Heald and Odum, 1970; Odum, 1971; Odum and Heald, 1972; 1975; and 
Odum fit fil-f 1972) . Briefly, the principal energy flow follow the path 

Mangrove Bacteria Detritus consumers Lower Higher 

leaf and (herbivores and * carnivores * carnivores 
detritus fungi omnivores) 

The chain begins with the production of carbohydrates and carbon 
by plants through photosynthesis. 


Leaf litter is then fragmented by the grazing action of amphipods 
and crabs (Head, 1971; Sasekumar, 1964) . Decomposition continues 
through microbial and fungal decay of leaf detritus (Pell et al., 1975; 
Cundell et al., 1979) and use and reuse of detrital particles (in the 
form of faecal material) by a variety of detritivores (Odum and Heald, 
1975) , beginning with very small sized invertebrates (meiofauna) and 
ending with such species as worms, molluscs, prawns and crabs, who in 
turn are preyed upon by lower carnivores. The food chain ends with 
higher carnivores such as large fish, birds of prey, wild cats or man 

The earlier findings have now been extended to include other 
energy and carbon sources to consumers in mangrove ecosystems, (e.g., 
Carter et al., 1973; Lugo and Snedalcer, 1974; 1975 and Pool et al., 
1975). In a recent appraisal of food chain dynamics, Odum et al. 
(1982) have enlarged the earlier basic trophic model to include inputs 
from phytoplankton, benthic algae and sea grasses, and root epiphytes. 
For example, phytoplankton may be important as an energy source in 
mangroves with large bodies of relative clear deep water. 

On this basis, the benthic algal contribution in estuaries with 
high levels of suspended sediments is likely to be lower. Similarly, 
where the continental shelf is truncated or very steep sloping, 
combined with high energy coastline and tidal amplitude, there is 
little sea grass or turtle grass. Where shading is not excessive, 
mangrove prop root epiphytes may also be highly productive. Values for 
periphyton production on prop roots of 0.14 and l.l gcal/m 2 /d have been 
reported. (Lugo et al. 1975; Hoffman and Dawes, 1980) . A generalized 
food web in mangrove ecosystem is depicted in Figure 2.14. 













Flf. 2.14: Generalized food web in manfrove ecosystem (after Burdtett, IMS) 


However, Odum et al. (1982) have stressed that in spite of 
innumerable studies, the Florida food chain model remains hypothetical 
and qualitative. Indeed, some recent data from the Indo-West Pacific 
region suggests that the Caribbean model requires some modifications, 

2.8.2 Primary wood production 

Estimates of total net primary production for mangrove forests 
are few because below-ground biomass is difficult to measure. The 
biomass, net productivity and annual litterfall of Rhizophora apiculata 
in the Ma tang, Peninsular Malaysia are shown in Table 2.9 together with 
figures from Phuket in South West Thailand. The biomass at 15 years in 
Thailand was less than in Malaysia, but productivity and the proportion 
of woody parts were similar. 

Wood is maintained as wood biomass, and what proportion enters 
the detrital pathway each year has not been determined. 

Table 2.9: Above ground biomass of Rhizophora apiculata 




Litter - 

Net productivity 

(t/ha/yr) Woody parts 

Matang, Malaysia 

Phuket. Thailand 

Source: Ong et al, 1980 - 400 m2 plots 

Christensen. 1978, 16 month observation of a 25 m2 plot. 

Lugo (1974) and Blasco (1984) record that measurements taken in 
some of the best mangroves indicate an average of 150 tons/ha of 
standing matter, which is quite low compared with other plant 
formations. Mangroves are therefore characterized by a particularly 
high productivity of organic matter in spite of a relatively low 
standing biomass. 

Termites and other organisms capable of breaking down the main 
components of wood (cellulose and lignin) , also called wood decay 
organisms, are important in nutrient recycling in the mangroves. The 
rate that slash decomposes due to these organisms also affects 
regeneration because slash not only hinders the distribution of 
propagules but impedes tidal flushing and reduces the amount of light 
available to seedlings. Furthermore, smaller plants are mechanically 
damaged by slash that moves with the tides. 


Litter fall is easily measured as a component of net primary 
production. Estimates of total litter fall in eleven locations world- 
wide range from 0.2 to 1.6 kg/m 2 /y dry matter (Woodroffe, 1982; 
Sasekumar and Lai, 1983). Litter fall estimates in the Indo-Pacific 
mangrove forests are usually higher. 


What happens to these leaves once they reach the forest floor? 
The Caribbean trophic model suggests that there is little accumulation 
of leaves in the forest as most of the leaf production is flushed out 
to channels and water bodies in the mangrove estuaries. In their 
riverine forest site Odum and Heald thus estimated that about 50 
percent of the mangrove litter was exported from the estuary (Odum et 
al., 1972) . 

In the Indo-West Pacific crabs are voracious feeders, and 
consume and/or bury a large proportion of the daily litter fall before 
it is washed away by the tides even in low tidal regions of the forest 
that are regularly flushed. Sasekumar and Lai (1983) estimated that 
crabs may consume or remove between 10 and 70 percent of the daily 
litter-fall before it is removed by the tides. Potentially, therefore, 
crabs are vectors of energy transfer between mangroves and forest 
sediments. Such rapid removal of litter by crabs, together with the 
potentially large amounts of wood production which may be processed 
within the forest, suggest that in contrast to many of the Caribbean 
mangrove systems that have been studied to-date (e.g Carter et al., 
1973; Lugo et al., 1975) , where there appears to be little retention of 
primary production (and hence nutrient stocks) within forests, primary 
and secondary production in the mangrove forests of the Indo-west 
Pacific region may be more closely linked. If this is correct, then 
pesticides which depress crab populations may cause a reduction in 
primary production in the long-term. 

As can be seen in Table 2.9, Ong et al., (1982) showed that in 
Matang, Peninsular Malaysia, ten years old Rhizophora apiculata 
produces 10.5 t/ha/year of litter compared to 6.9 t/h/year at 5 years. 
This is biologically and ecologically important because of the 
importance of litter fall in the food web. Forests should ideally be 
managed not only to optimize wood production but also to sustain an 
active ground meiofauna (microbial) and macrofuana. Environmentally 
therefore, well managed forests should promote vigorous growth with 
high biomass increments capable of producing high levels of litter 
fall. In this context, the impact of pesticides and herbicides on the 
meio-macrofauna should also be considered, particularly in habitats 
where bivalves (Anadara sp) abound. In the Sierpe-Terraba area in 
. Costa Rica, copper-based contaminants from the banana plantations have 
significantly reduced the once thriving molluscan population. 

2.8.3 Secondary production 

A wide range of food-energy resources is available for secondary 
production in the mangrove ecosystem, including the primary 
productivity of phytoplankton in the photic zone, the production by 
benthic macrophytes, and the particulate and soluble organics that 
originate from the productive mangrove forests. Ong (1985) estimated 
that the primary productivity contributed by the mangrove forests to 
the Straits of Malacca is about 0.22 g/m 2 /day, a figure at least equal 
to that of the phytoplankton productivity of the Straits of Malacca. 

The importance of the primary productivity of mangrove forests as 
a source of food for especially aquatic resources is further supported 
by the series of studies conducted in the Selangor mangroves (Sasekumar 
et al., 1984; Thong and Sasekumar 1980). 


Leh and Sasekumar (1984) found that various inshore penaeid 
species consumed 12i-36% of plant matter, of which ll%-59* was 
identified to be of mangrove origin. Mangrove detritus formed 32%-42% 
of the gut content of the planktonic shrimp Acetes 8p (Tan, 1977) . 
This shrimp is a key linkage organism because the adult white prawns 
Penaeus tnerguiensis feed heavily on Acetes sp and mysid shrimps (Chong 
and Sasekumar, 1981) . Several omnivorous fish species likewise feed on 
Acetes sp as well as on mangrove detritus (Ong, 1977) . 

2.8.4 Keystone species 

In the coastal ecosystem certain wildlife species play pivotal 
roles in maintaining an equilibrium in the fauna. If these keystone 
species/ as for example a predator is removed, the community will be 
thrown out of equilibrium. Crocodilus porasus is one such keystone 
predator in the Sundarbans. Crocodiles feed largely on less commercial 
fish, some of which at some stage prey on the more valuable fish and 
fish eggs (Bustard, 1975) . Their movement also activate the movement 
of nutrients in the water and is a keystone in maintaining a productive 
fishery (Box 2.3) . 

The role of catenas 

"Fishermen in the Amazon had for some time noticed that wherever caimans 
disappeared/ fishing declined. This decline was rather unexpected, fishes being 
the main food of the caimans. It seemed to the natives that in the absence of 
the predators, fish would multiply. It is now clear from research undertaken in 
Amazonian mouth-lakes that this is the reverse of the truth. Biomass in tropical 
aquatic systems consists almost exclusively of animals {whereas the fauna 
represents only a small fraction in rain forests). Umnotogist E. J. Fittkau has 
researched the amounts of nutrients released by caimans in the course of their 
metabolism and the impact of these nutrients on primary production. 
Measurements proved that caimans daily add nutrients {mostly of allochthonous 
origin) in sufficient quantities to effect an increase of primary production and an 
attendant enlargement of the autochthonous food chain/ 

Source: Prof. Federico Medam, 1976. 

Box 23: Caimans as keystone species 

Foresters and other ecologists are also familiar with the role of 
monkeys (Macaca mulatta) and deer (Axis axis) where a significant 
portion of the food of deer may come from food particles dropped by 
foraging monkeys. 

Yet another example is found in the Sundarbans, where, if the 
tiger population is reduced, both deer and wild pig (Sus scrofa) would 
increase to the point of reducing forest regeneration and damaging the 
forest. In Malaysia, bats, which shelter in mangroves, are important 
in pollinating durian flowers. 

In Australia, birds feeding on sweet flora exudate of Rhizophora 
spp also reduce the insects that destroy terminal buds and shoots 
(Primark and Tomlison, 1978) . 


2.8.5 Management implications 

As our knowledge on trophic relationships and interactions 
improves, foresters will be able to manage their resources better 
without harming the environment. Figurt 2.15 shows some of the 
positive and negative impacts associated with the use of mangrove areas 
under production forestry, open- water aquaculture, capture fishery and 
conversion to agriculture and pond -culture. 



^ Ni 

1 OVtnFISHWO It ' 


\\_ , 


EES j \ 

/ .mS5t 


. . ,, ft. . j|l. 1 *WWMMM 



// FRESHWATER OUTROW \\ 4 ? 1Airfv 
*WNDW*AK // \\ JWJJ" IV 

ft \^ 1 



Fig 2.15: Some ecological interactions between various land uses and economic activities 

in mangrove areas. (FAO Environment Paper No. 1, 1982). 

In mangrove management it is thus essential to take a holistic 
approach and to secure the survival of the entire ecosystem. 
Conserving or promoting biodiversity through the selection of species 
to be felled and regenerated and the protection of habitats for various 
marine and terrestrial animals is an imperative as is the maintenance 
of the protective role the mangroves play along river banks and 

Riverine vegetation should therefore never be felled 
indiscriminately, as bank erosion will increase water turbidity and 
adversely affect aquatic fauna, particularly shrimp larvae, molluscs 
and the breeding of important estuarine species. Protective areas 
should also be set aside in the mangrove area proper for the 
conservation of wildlife and plants of special interest. 

Where the demand for land for agriculture or aquaculture 
necessitates the conversion of mangrove areas, the sites should be 
properly evaluated prior to the conversion in order to minimize the 
damage to the mangrove ecosystem as a whole. 



This Part highlights the multiple-use potential of mangroves as a 
renewable resource. It gives a coverage of selected major wood and non-wood 
uses as well as their interrelationships. 


To be conserved, a resource must be managed sustainably and seen to be 
useful to local communities. In this context, extractive activities should 
produce a positive impact on the surrounding community - such as generating 
local employment - without impairing the environment. It is almost impossible 
to conserve a resource without the support of the local population. Local 
participation may involve information sharing, consultation, decision-making 
and at the highest intensity initiating action. 

Given the multiple use potential of mangrove ecosystems, an integrated 
approach to mangrove management is essential and should cover the full range 
of products and services which can be obtained from these areas. 


The uses and values of the products obtainable from mangroves are many 
and important. The importance of the resource stems from the many products 
taken directly from the mangroves, including the non-wood products, as well 
as amenities provided from within and beyond its boundaries. Wood products 
range from timber, poles and posts to firewood, charcoal and tannin. Non-wood 
products include thatch, honey, wildlife, fish, fodder and medicine. In 
addition, mangrove lands are often converted to salt ponds or to agriculture 
or aquaculture purposes . 

Many of the non- timber species found in mangroves are extremely 
versatile. The Nipa palm (Nypa fruticans) for example, is used mainly as 
thatch for roofing but can also produce a sugary syrup, alcohol and vinegar. 
Phoenix paludosa palm stems are used in fencing and construction purposes. 
Nibong (Oncosperma filamentosa) is a very useful palm in the landward fringe 
of mangroves. Whole stems are used for house and bridge posts and split stems 
for flooring, decking, fish drying platforms, backbones for nipa thatch, roof 
gutters, water pipes, and many other uses. 

Other plants are used as fodder. Avicennia leaves, for instance, are 
grazed by camels, goats and cattle in India, Pakistan, and the Arabian coast. 
In Australia, wild buffaloes graze on mangroves in the Northern Territory. 
This sight can also be seen in Vietnam. The stall feeding of sheep and pigs 
has been practised in a number of countries using mangrove fodder in 
conjunction with other feedstock. 

Cockles (Anadara sp) collected from estuarine mud-flats are a source of 
protein for local inhabitants, as are crabs and the fishery in the tidal 
aquatic system. 


A list of some mangrove-based products is shown in Table 3.1. 
Table 3.1: Mangrove-based products 

Sweetmeats (propagules) 
Vegetables (fruit/leaves) 

Household items 


Hairdressing oil 
Tool handles 
Rice mortar 

Match sticks 



Paper products 

Paper - various 

Other products 

Packing boxes 
Wood for smoking 

sheet rubber 
Fuelwood for:- 

salt making 

brick kilns 


tobacco drying 

B. Other Natural Products 






Reptiles/Other fauna 

Adapted from UNEP, 1983 

A. Mangrove Forest Products 




Timber, scaffolds 
Heavy construction 
Railway sleepers 
Mining props 
Boat building 
Dock pilings 
Beams and poles 
Flooring, panelling 
Thatch or matting 
Fence posts, chipboards 


Fishing stakes 
Fishing boats 
Wood for smoking fish 
Tannin for net/lines 
Fish attracting shelters 

Textile, leather 

Synthetic fibres (rayon) 
Dye for cloth 
Tannin for leather 

Food, drugs & bevaragas 



Cooking oil 


Tea substitute 

Fermented drinks 

Dessert topping 

Condiments (bark) 

Among the "intangible 11 benefits, often taken for granted, are: (a) 
coastal protection against wave and wind erosion; (b) moderating the effects 
of coastal storms and cyclones; (c) shelter and habitat for diverse wildlife, 
particularly avifauna; (d) nutrient sink-effect and reduction in excessive 
amounts of pollutants, and (e) entrapment of upland runoff sediments thus 
protecting nearshore reefs and reducing water turbidity. Mangroves also 
provide opportunities for education, scientific research, recreation and 



Mangrove forests have favourable silvicultural characteristics which 
lend themselves to intensive forest management for woody products. Some of 
these characteristics are as follows: 

Rapid growth: Mature stands under suitable conditions 
may yield over 270 m 3 /ha within 30 years, equivalent to 
an MAI of 9-10 m3/ha; 

Good regeneration potential: Most mangrove species 
flower and fruit regularly and the propagules are 
dispersed by tides. Thus, mangrove stands can recover 
rapidly from natural or man-made disturbances, including 
intensive logging. 

Tendency to form hooogeneoui/even-aged stands: Pure 
stands of Rhizophoras or Aviceiwias are not uncommon and 
even in mixed stands, the principal components are 
restricted to a handful of species; 

Diversity of forest products: A wide range of products 
are produced and as bioenergy plantations even the 
smaller thinnings may be used as firewood. 

Environmental, biological and social considerations require that 
harvesting operations must not impair the ecosystem. In particular, wood and 
non-wood removals should avoid the following: 

deny indigenous communities of their traditional access 
to a reasonable harvest of mangrove products 
significantly alter substrate composition; 
alter the local patterns of surface water circulation; 
exceed the biological productive potential of the site; 
reduce the regeneration potential of desired species; 
reduce the protective functions of the forest; or, 
damage wildlife breeding, nursery or shelter sites. 

Large scale capital-intensive operations are not advisable. For this 
reason the clear-felling of large mangrove areas for exporting chips to 
developed countries may be detrimental to local ecologies and economies in the 
long term, although for a time it may generate some foreign income. 

Although a clear-felling system has been practised in the Matang 
mangroves for almost a century without apparent adverse impact upon the 
environment and ecosystem, this is the exception rather than the rule. 
Blessed with a buoyant economy, abundant fuelwood supply, natural petroleum 
and gas, and guided by a sound land use policy, the pressure on mangroves in 
Peninsular Malaysia is less than that in most other countries. 


More importantly, individual felling areas are relatively small and 
sufficient funds are provided for reforestation. However, when clear-felling 
systems are applied to other situations without reference to local ecology and 
the country's socioeconomic context, the mangrove resource can be destroyed 
or degraded very rapidly. 

3.2,1 Timber 

Under favourable conditions, mangrove trees can grow to large 
sizes. Rhizophoras over 40 m tall are not uncommon and individuals 
over 62.5 m have been reported (Sukardjo, 1978). However, large trees 
are becoming scarce, especially in South East Asia, as most of them are 
removed before they can attain such sizes. 

Rhizophora spp are however, not valuable as timber because of 
their tendency to split and warp when dried. The wood is dense and 
difficult to work. The sapwood 4 is easy to preserve but not the 
hardwood. It is resistant to decay but not to marine borers. Its 
possible uses include agricultural implements, boat construction (knees 
and ribs), general heavy construction (rafters, beams, joists), marine 
and bridge construction (underwater, non- teredo infested waters) , 
marine and bridge construction (above water), fence posts and poles. 
In the Ca Mau peninsula in Southern Vietnam, a small amount of 
Rhizophora is used for walling and flooring. Sugden and Von Cube 
(1978) also indicate that R. racemosa can be used for a variety of 
products including particle board, railway ties and posts. 

The wood of R. mangle is exceedingly heavy with a specific 
gravity varying from 0.9-1.2. It is comparable in density to 
Greenheart (Ocotea rodiaei) and Bui let wood (Manilkara bidentata) . The 
green weight was determined to be 1 200 gm/m3 (75 Ibs/cu.ft.) at 46% 
m.c. (FAO) . Avicennia germinans, which has a lower density (about 
0.64) and good nail holding qualities, is often used as railway ties in 
Cuba. In Venezuela A. nitida is used as mining props, telegraph and 
transmission poles. Table 3.2 shows the densities of selected species. 

TaMe3J: Wood density for selected spedcs 


Wood EfeMity . | 

Oven- dry 


Air-dry I 

Avlcermia nltlda 






Rhizophora harrisonii 



1.045 1 



Rhizophora mangle 






Arroyo, 1971 (Figura* within brackets dtnota 

content %) 

In the Bangladesh Sundarbans, Heritiera femes (Sundri) is the 
prime timber species used for house and boat construction, while the 
tops are used for hardboard and as firewood. Creosoted Bruguiera 
gymnorhiza transmission and telegraph poles were used in the Andaman 


3.2,2 Charcoal 

The energy given off from a piece of wood is about the same on a 
weight basis irrespective of species (Openshaw, 1983) , Wood density 
largely determines charcoal yield, consequently a given volume of wood 
will result in different yields of charcoal (measured on a weight 
basis) dependent on the species. For example, a cubic metre of air 
dried wood (15% m.c.) will give the following approximate weight of 
charcoal including fines for various species. 

Table 3 J: Yield of charcoal for selected species 



pical Hardw 





tfaight of charcoal 
in kg/*3. 





For bioenergy plantations, the potential energy yield per unit 
area (i.e. above-ground Jbiomass) rather than volume is the important 
measure. For example, a pine plantation with twice the standing volume 
of a Rhizophora forest has 24% less potential energy per unit area. 
The high heat value of Rhizophoras varies between species, but a value 
of 4 400 kcal/kg has been cited for R. mangle in Ecuador (Doat, 1977). 

Rhizophoras are preferred for charcoal making. Their moisture 
content (MC) when felled is about 40% (as % of oven dry weight) 
compared to Avicennia wood which ranges from 70-95%. JUiizpphora wood 
dries to about 25% MC after two months, whereas Avicennia requires up 
to six months to dry to 35% MC. This partly explains the popularity of 
Rhizophora wood, as predrying stock can be kept to a minimum. Charcoal 
outturn is improved when dry billets are used because less energy is 
needed to dry the wood. 

Other species (Bruguiera gymnorhiza and Ceriops sp) are also used 
but in smaller quantities. 

Charcoal is the main mangrove product in Thailand, Peninsular 
Malaysia, Sumatra (Indonesia) , Myanmar and Southern Vietnam. 
Industries are well developed at the village and cottage industry 
levels in most Asian countries where mangroves still abound. Charcoal 
is mainly used for cooking purposes and small-scaled industries. 
However, in West Africa firewood is more commonly used. 

In Matang charcoal is produced in dome-shaped, masonry kilns. 
These are located along small rivers or creeks to facilitate transport 
of billets. The battery of kilns is covered with Nipa roofs. The 
roofing requires little attention as the tar- laden smoke emitted by the 
kilns preserves the Nipa thatch. However, if the kilns are not fired 
regularly, the masonry structure and Nipa thatching deteriorate 
rapidly. Masonry kilns are long-term, location specific and costly to 
construct. To be economically viable there must be an assured supply 
of billets and reasonably low land costs. In contrast, earth pits are 
easy to build, costs are low and the structures are often temporary. 


Freshly cut billets are normally debarked in the forest. Billets 
of 18-23 cm diameter and 1.6 m in length are stacked vertically inside 
the kiln. High grade charcoal of uniform quality is made when air- 
dried, debarked billets of uniform sizes, densities and species are 
used. Bricks are used to support the standing billets for even 
burning. This reduces the amount of partially carbonized ends or 
" brands" produced. Where Rhizophoras are limited in supply, inferior 
species such as Avicennias and rubber wood (Hevea Jbrasiiensis) are 
substituted for kindling wood. 

The conversion efficiency is still far from efficient. In 
Ma tang, a standard 6.7 m diameter dome-shaped kiln operates at only 19% 
efficiency. About 55 t of greenwood per kiln is required for an 
efficient burn (Table 3.4) . Current use of smaller and lower density 
billets compared to wood harvested from the virgin stands may partly 
account for the reduced conversion efficiency in Matang. (Harun, 1981) . 

Table 3.4: Greenwood input and charcoal output per bum (tonne) 



Greenwood Required 


Charcoal produced 

















Port ttold 

















Kuala Trong 

















Sungei Karang 

























Source: Watang Forest Reeerve Working Plan (1980-89) 

Frisk (1984) estimated that a kiln of 6.7 m diameter and 7 m in 
height requires 9 t of clay, 9 t of fine sand and 15 000 - 17 000 
pieces of bricks (6 cm x 11 cm x 23 cm) to construct. It costs about 
$7 000 including the roof shelter. 

In Yeesarn, Thailand, the estimated cost, including Nipa thatch 
roofing, was $2 284 for kilns of 5-6 m diameter and 3 m high. Kilns 
last a long time provided worn out, porous or broken bricks, caused by 
thermal cyclical stress and normal wear and tear, are regularly 
replaced. Cracks should also be sealed immediately. 

Charcoal making can be profitable, but it is less lucrative than 
shrimp -farming. In Matang, a hectare of mangrove managed for charcoal 
yield the government a net revenue - in form of royalty, premia, 
license fee, fines etc. - of US$ 478 (Othman and Khan, 1984) . The 
market value of the same forest is very much higher at US$ 8 333/ha. 

In West Africa, Central America and the Caribbean Islands, 
charcoal is mostly made by the earth pit or mound method. Generally, 
these are less efficient, and produce charcoal of variable quality. 
Greater care is also required in tending and controlling the 
carbonization process. 


Figure 3.1: Masonry kiln for charcoal making in Indonesia. 

Photo by M.L.Wilkie 

figure 3.2: Charcoal making by the earth mound method in Guana), Cuba. 

Photo by P.W.Chong 


In Cuba, billets of all sizes and lengths are arranged vertically 
to form a large circular stack. This is then covered with fern fronds 
and sand and sealed with mud. 

In Costa Rica, the "carboneros" construct their charcoal -pits 
along the beach above the normal high tide level. The dimension varies 
from 3 to 27 m in length, 0.3 m in depth and 1.1 - 2.1 m in width. The 
pits are oriented perpendicular to the shoreline, so that billets can 
easily be rolled into the trenches to form a stack about 0.9m high. 

Running along the length of the trench are two rows of supporting 
billets to facilitate the positioning of the billets and to improve air 
circulation. The billets vary in length from 1.1 - 1.5 m. and are not 
debarked. The fully charged stack is partly above ground level. It is 
then covered and sealed with Acrostichum leaves, earth and sand, and a 
plastic cover to keep out the rain. One end is fired, and 
carbonization progresses along the stack at the rate of about a metre 
per day. This is monitored and controlled by several crudely made 
vents. The charcoal produced is of variable quality, mixed with sand. 

A conversion efficiency of about 13% is achieved, equivalent to 
an input /output ratio of 1 : 7.7, based on dry weight. Higher output 
is possible with pre-dried billets but the added costs of holding stock 
must be considered. For masonry kilns the conversion efficiency is 
around 19% to 22%. Higher conversion efficiency means that less wood 
needs to be grown, harvested, pre-dried, transported and used for 
charcoal making. 

Unplanned promotion of charcoal industries, without proper 
resource assessment and forest management, will, however, only hasten 
the ecological destruction and depletion of the mangroves. To avoid 
this, the construction of new or enlargement of existing kilns and 
earth-pits should be monitored and regulated by the responsible 
forestry authorities. Similarly, only boats which are registered with 
the Forest Department should be permitted to carry forest produce to 
minimize illegal felling. 

Agricultural activities are essentially seasonal in nature, 
whereas, forestry activities can be carried out throughout the year. 
Thus, forestry activities such as charcoal making can offef employment 
opportunities to counter seasonal unemployment. 

Well organized charcoal industries linked to sustainable supplies 
derived from regulated State forests and/or private plantations and 
woodlots can contribute significantly towards rural employment, economy 
and rural industrialization. Where the commodity is produced 
efficiently and marketed competitively, it can be transported over vast 
distances to serve the needs of urban and rural consumers alike. 
Surplus charcoal may even be exported to neighbouring countries. 
Furthermore, charcoal production can be strategically planned to 
complement and support a country's rural coastal dendro-energy plan 
within the national energy framework thereby reducing its over- 
dependence on non-renewable fossil fuels. 

It should be emphasized that the market for charcoal varies 
widely between and within regions. 


Southeast Asia has always had a large internal market and a 
lucrative export market that supports large charcoal industries. In 
turn, this has made the management of the mangrove forests in Malaysia 
and recently also in Thailand economically viable. In tropical 
Ameraica however, the population does not have the same tradition for 
cooking with charcoal and investments in the management of ttuw^rtivies 
can today rarely be justified by charcoal production alone. Mangrove 
regions in Africa lies in general between the above regions in respect 
of conditions of charcoal markets and hence the prospects of basing the 
economy of mangrove management on charcoal production. 

Access to credit and finance is an important institutional 
requirement as improved carbonization methods require substantial 
capital investment, organization and training of operators. 

3.2.3 Firewood 

Rhizophoras are favoured as fuelwood for domestic purposes and 
are commercially removed as in the case of Matang and Thailand, or 
collected by fishermen and villagers. 

In Sierra Leone, large quantities of Rhizophora racemosa firewood 
are used for fish smoking (banda process) . Since the weight of 
firewood consumed in processing more than equals the throughput of 
fish, it is not surprising that cutting of mangroves for fuelwood has 
become a major occupation within the fishing community. Table 3.5 
below on fuelwood/fresh fish weight ratios for smoking fish using the 
"banda n method is based on studies conducted at Yelibuya Island. 
(Chong, 1989) . 

Table 3.5: Fudwood/fish ratios in fish-smoking - Sierra Leone 

Fresh Dry 
Fuel ratio Sample size 

Hard Dry 
Fuel ratio Sample size 

Source: Seymour. T; 1987, KFDP Technical Report 6/87, 

Fuelwood is also used .in boiling brine to produce salt. The 
heating value of selected mangrove species are shown in Table 3.6 

Table 3.6: Heating value of selected raanfro ve sped* 

Species Cal/gw | 

Rhizophora apiculata 
Ceriops tagal 
Sooner at i a alba 
Brugui'era par vi flora 
Avkennla offlcinalls 
Xylocarpus granatun 



Figure 3.3: Rhizophora firewood bundled with split mangrove stilt-roots. 

Sierra Leone. Photo by P.W.Chong 

Figure 3.4: Bark collection in southern Vietnam. 
Photo by P.W.Chong 


3.2.4 Fishing stakes /poles 

In Singapore, Hongkong and Malaysia, there is an established 
demand for mangrove piling poles used in land reclamation and the 
construction industry. Used in wet sites which are not infested by 
shipworms, such mangrove piles can outlast non- treated inland 
hardwoods. Currently poles are imported from the Indonesian islands 
and from southern Thailand. 

Along the coastal waters, Nibong (Oncosperma filamentosa) is 
normally used as fishing stakes in Southeast Asia, but sometimes 
mangrove poles are also used. These have to be regularly replaced. 

Along the muddy river banks, small fishing stakes are used to 
support tidal fish nets. Mangrove poles are also used for scissor nets 
in housing construction. In the Cuban waters and in countries in South 
East Asia, fishermen cut mangroves and dump them into the shallow 
coastal waters as a way of creating shade and thus attract fish (fish 
attraction devices) . 

3.2.5 Pulp 

Gewa (Excoecaria agallocha) is the principal pulping species used 
in the newsprint mill in Bangladesh. Sonneratia caaeolaris, Excoecaria 
agallocha, and Avicennia marina produce strong Sulphate pulps. 

Large mangrove concessions have been granted for chipping 
operations in the east Malaysian states of Sabah and Sarawak and in 
Indonesian Kalimantan and Saluwesi based on Rhizophoras and Brugruieras. 
The chips are exported mainly to Japan for making dissolved pulp and 
cellulose derivatives such as rayon, used in the textile industry. The 
African species of JR. raceznosa is reported as suitable for making 
dissolved pulp although some problems exist due to the inorganic 
crystals present in the wood (Sugden and von Cube, 1978) . 

The production and export of chips from mangroves have in parts 
of Indonesia led to the clearfelling of large areas in one harvest. 
The regeneration of these areas has proven unsuccessful in many cases 
and it is now recognised that harvest areas should be kept small and 
scattered if satisfactory regeneration is to be obtained. 

3.2.6 Tannin 

Rhizophora bark produces very fine tannin of the 
phlobaphene-yielding catechol group which is not broken down by 
ferments and is thus very suitable for leather work. Tannin from 
mangrove species has also been used for curing and dyeing of fishing 
nets made of natural fibre to make the nets more resistant to 
biological decay. 

The amount of tannin varies somewhat with bark thickness, 
position on tree stem, location of tree, dryness of the bark, as well 
as between species. The bark must be fresh and transported to the 
tanneries as soon as possible in a moist condition. Ceriops 
candolleana, Carapa obovata and Rhizohpora inucronata were at one time 
the favoured tan-bark species exploited in Malaysia. 


Although the number of tinteros (bark collectors) is small in 
Panama, Costa Rica and some Latin American countries, their impact on 
the mangroves may be much greater than felling activities undertaken by 
other groups because commercial quality bark is harvested from only the 
largest and better Rhizophora trees with straight boles. When these 
dominant stems are felled, considerable damage is caused and 
innumerable gaps are created in the forest canopy. The debarked stems 
are left to rot in the forest thereby creating large amounts of debris 
that encourage termite infestation in the drier sites. It takes about 
2 year for the smaller slash to decay but a much longer time is 
required for the large stems to decompose. 

In Costa Rica because of the high regeneration potential in the 
Sierpe-Terraba mangroves, these gaps revegetate quickly but not 
necessarily with the desired species, since P. rhizophorae, which is 
not exploited, may dominate these gaps. 

The production of tannin has declined greatly in recent years, in 
particular since local demands have been reduced after the introduction 
of nylon fishing nets and the use of chrome as the predominant agent 
for leather curing. 

In Colombia, the extraction of bark for tannin has been 
prohibited several years ago because of over-exploitation of the 
mangroves on the Pacific coast. Today, mangrove exploitation for wood 
and bark is totally forbidden in this country. 


The traditional "management paradigm 11 implies that if the forests are 
properly managed then the non-wood ecosystem components will, ipso facto, 
remain stable. This is notionally flawed, because unless the non-wood 
components are integrated into the planning, implementation, and monitoring 
levels of the forest management system adopted, they will often be 
marginalized or ignored. 

The inland margin of the mangroves and the upper tidal limits of 
estuaries constitute the brackish water zone, where the water is mildly 
saline. Economically, this is an important zone because, the brackish-water 
creeks are fringed by the salt water palm, Nypa fruticans. The mangrove Date 
Palm, Phoenix paludosa and Sago (Metroxylon sagu) are also found in this 
formation. The inhabitants of the mangrove- Nypa palm zone along the Gulf of 
Papua New Guinea subsist almost entirely on a diet of sago, which is very rich 
in carbohydrates, and crabs, as a protein source. 

Mangrove forests are the habitat of numerous species of fish and 
shellfish. Coastal fisheries depend on these and they provide much of the 
protein needed by coastal people. Some mangroves are converted to fish- or 
shrimp-ponds. Near urban centres, much mangrove forest has been lost to 
provide land for industry, tin-mining, solar salt-works and for hotels. There 
has also been conversion to paddy fields, often unsustainable because of acid 
sulphate conditions. A review of selected mangrove -based, non-wood resources 
including minor forest products, fishery and conversion of mangrove land to 
other land uses is given in the following paragraphs. 

3.3.1 Nipa palm 

The uses of this palm are many and diverse. It yields an 
important thatching material, which is used for the roofs and walls of 
rural houses. The villagers cut the fronds leaving behind about 2-3 
younger fronds per culm. The leaflets are removed from the main stalk 
and usually soaked in salt water for several days to soften them. The 
leaflets are then folded over a spine of split Nibong (Oncosperma 
filamentosa) 4-6 feet long, stitched in place with fine split rattan, 
and dried in the sun. 

The completed shingles (in Malaysia known as "atap") are made in 
several qualities. They are cheap, light to transport, easy to fix and 
can last several years, particularly when used in houses with open 
stoves. One of the unintended effects when improved cooking stoves 
with chimneys are introduced into rural households is that their atap 
roof will not last as long! Cigarette wrappers are also made from the 
young shoots of Nipa. 

Another potential of the Nipa lies in the sugary sap of the 
flower stalk, which can be used to produce a sugary fluid. As the 
removal of mature fronds will reduce the sap yield, Nipa plantations 
managed for alcohol production cannot be harvested for thatch making. 
In the Philippines the cultivation of Nipa for alcohol production has 
been practised on a considerable scale for many years. 

Plantation establishment 

The seedlings are ready for transplanting when they are about 18 H 
high. About 450-500 palms are planted per hectare. Split bamboo 
guards are used to protect the shoots of the young plants from being 
destroyed by crabs. Nipa fully matures within 5-6 years but bears 
fruits after three years or earlier. 

Extraction of sap 

The fruits, borne in a cluster on a large flower stalk, take 
about three months to develop. When the head is well developed, but 
before the skin of the fruits begins to darken in colour and becomes 
hard, the stalk is massaged. 

This process known as "goncbang* involves swaying the stalk, 
gently at first but with increasing severity for about 3 weeks, at the 
end of which time it is violently shaken. The fruit head is severed, 
and the exudate is collected. A thin slice is made every day or even 
twice a day and the stalk will continue to yield for several months. 
As many as 26 fruiting heads at various developmental stages have been 
observed on a single palm. 

In Malaysia, tapping is undertaken all year round without any 
apparent ill effect. Reportedly, each spathe daily produces about 0.49 
litres of sap. Two spathes per palm can be tapped continuously to 
yield annually about 252 litres based on 260 working days. In Papua 
New Guinea, a mature palm produces about 200 litres per year. With a 
plantation of 250 palms/ha, the annual sap yield will be some 50 000 


Figure 3.5: Collection of typa leaves in the Sundarbans, Bangladesh 

Photo by M.L.Wilkie 

Figure 3.6: SWngJes (Atap) made from Nypa leaves in Sumatra, Indonesia. 

Photo by M.L.Wilkie 


The sugar content, mainly sucrose, varies from 6-17 percent. A 
conservative estimate of only 6 percent sugar content would yield some 
3 000 kg/ha/year of sugar or about 5 400 1/ha/year of alcohol. To be 
economically viable, large scale Nipa plantations are needed. 
Additionally, as the juice ferments very rapidly, an efficient and 
rapid juice collection system is very important. 

Apart from alcohol, three alternative products might also be 
profitably made from Nipa sap. The simplest product is sugar syrup. 
which can be marketed as a speciality sweetener like maple syrup. This 
is noteworthy because in 1986 maple syrup production plummeted and 
retail prices rose to nearly US$ 50/gallon. Another potential product 
is brown sugar, which is popular in developed countries as a form of 
"health" food. The third and possibly most important product is 
vinegar, which can be used in domestic cooking, industry and for 
preserving food. Nipa vinegar could be an important substitute for 
industrially produced vinegar. 

3.3.2 Apiculture 

Honeybees, from the genus Apis, have been beneficially and 
destructively exploited by man for thousands of years. Apis meiiifera, 
which is native throughout Africa, most of Europe and the Middle East, 
is the best known and most widely spread species. 

In Africa, honey is 
still being collected from 
wild nests, but the common 
types of traditional 
beekeeping use hollowed log, 
bark, basket or clay hives 
placed on tree branches. In 
West Africa, the nectar and 
honey and pollen potential 
of the mangroves has yet to 
be fully exploited. (See 
Figure 3.7) 

Fig. 3.7: Bark hive placed on tree in Africa 

There are no honeybees native to the Americas, Australia or the 
Pacific area although during the last 400 years or so Apis mellifera 
has been introduced from Europe to these areas (Bradbear, 1990) . 

Avicennia germinans, Laguncularia racemosa and Conocarpus erectus 
are important sources of nectar and pollen in these areas, and 
Rhizophora mangle is also reported to be a melliferous plant (Hamilton 
and Snedaker, 1984) . Off-season flowering plants enable bees to build 
up sufficient stores to survive during stressed periods such as cold 
weather, drought, monsoon rains and floods, or devastating bush fires. 

Within the mangroves modern beekeeping with comb-hives are 
practised in Florida, the United States and Cuba. About 40 000 hives 
are relocated to the fringes of mangroves in Cuba between April to June 
when pollen and nectar production from inland natural vegetation and 
crops are very much reduced. 


Apiculture in Cuba is based on local hybrid (Apis mellifera 
mellifera x Apis mellifera lingustica) . About 25% of the annual total 
honey production in Cuba (some 8-10 000 t) is derived from the 
mangrove sources. 

In Asia, apiculture is an important activity in Burma, Bangladesh 
and India (1984, loc.cit.). At least three honeybees are native to 
Asia and all are exploited by man. Two of these, the Little honeybee 
(Apis florea) , and the Giant honeybee (Apis dorsata) , cannot be kept in 
hives as they nest in the open air, on a single comb. The former 
builds its small comb (about 25 cm diameter) hanging from branches 
within bushes, while the latter suspends its much larger combs (around 
1 m in diameter) from tree branches, rocky ledges and buildings. The 
Giant honeybee's nest may well contain 50 kg of honey. The third 
species, Apis cerana, is known as the Asian hive bee and can be kept in 
a hive (Bradbear, 1990) . 

Honey production depends on the type of bees, the availability of 
pollen and " nectar, prevailing wind, temperature, salinity, 
contaminants, availability of freshwater and other factors. Aerial 
spraying of pesticides can seriously affect beekeeping. Seasonal 
burning of the inter- terrestrial zone and the mangrove landward fringe 
will also destroy a number of melliferous plants. 

In the Sundarbans, beeswax and honey are produced by wild bee 
swarms that build hives on branches, in tree holes and crevices. The 
hives and trees are often destroyed during collection. It was 
estimated that about 9 300 trees were felled in the 1982/83 season to 
produce 233 tonnes of honey and 58 tonnes of beeswax, whereas under 
proper management about 1 550 hives would have sufficed (Christensen 
and Snedaker, 1984) . It has been suggested that the setting up of top- 
bars suitably baited with swarm attractants could reduce the number of 
trees felled or damaged. 

Beekeeping fits in well as part of integrated rural development 
programmes, and the best projects are those which promote sustainable 
beekeeping on a long-term basis making use of indigenous expertise, 
knowledge and materials. Imported equipment should be avoided as 
supplies may be unpredictable, or may later be obsolete because of lack 
of spare parts or suitably trained maintenance technicians. Some 
beekeeping management objectives are given below: 

To alleviate rural poverty by creating in-s/tu income generating 

activities through beekeeping; 

To improva the potential for beekeeping by planting melliferous 

mangrove spades toward the landward fringe; 

To improva the quantity and quality of beekeeping products through 

sound management; 

To assist in the making of beekeeping products; 

To overcome specific problems, e.g. disaasa or pesticide misuse; 

To transform destructive honey-hunting from wild nests to sustainable 



Some points on good bee colony management practices are 
summarized in Box 3.1 below: 

(!) Use food fcNcs that ire mtde from toed, (teabte and iaexpeflrive 
materials that ate suited to the biological roprircsnents of bees, 
penottthq; their fall development; 

(2) Uie belted htm and locate tern dole to mefflftroui ptote. Bee* 
wax, propolis or other materials are used as baits to entice bee swarms; 

(3) Access to freshwater must be provided dose to hives; 

(4) Him should be ventilated and wifl-protected against theft and 

(5) Beekeeper mast use protective clothing and equipment; 

(6) Bees should be regularly inspected; 
(?) Use good strains of bees; 

(8) Honey and bees-wax must be harvested in time; 

(9) The hooey and bees-wax must be separated; 

(10) The honey shotdd be stored to a coot place and the btt*~wax 

Source: Ntenga, G.M and Mugongo, B.T, 
Box 3.1: Notes on good bee colony management 

3.3.3 Wildlife 

As in other forest types, the wildlife in the mangroves is an 
important source of protein for the local community. In addition, some 
species, especially reptiles, are hunted or reared for their hides. 
Examples of traditional utilization of selected wildlife species found 
in mangroves are described in the following: 

The rodent Hutias (Capromys sp.) , that live in the mangroves, are 
endemic to Cuba, comprising C. sanfelipensis , C. garridoi, C, 
angelcabrerais, C. auritis and C. oilorides. The meat is highly 
relished by the local people. 

The giant forest hog (Hylochoerus meinertzhageni rimator) is 
often found marauding in the swamp margin and it is a source of bush 
meat to the West Africans, similar to the wild boar (Sus scrofa) in 

In Central America, the large lizards, Iguana iguana (iguana) and 
Ctenosaura si/nilis (garrobo) are found in the mangroves, the former is 
quite common. These two out of six species present are most commonly 
eaten by the local people. Hunting pressure does not seem to have 
diminished their population. Ctenosaurs may be amenable to large-scale 
captive breeding as a pile of cinder blocks can support flourishing 
populations! On the other hand, iguanas have been virtually 
exterminated in some Central American areas (e.g., El Salvador). 

The West African Monitor Lizzard (Varanus exanthema ticus) is also 
hunted for food. 


In Costa Rica, the Pacific 
Ridley Turtle (Lepidochelys 
olivacea) is relished for its meat 
and weighs an average of 40 kgm. 
Its numbers are diminishing 
because of predation and over- 
exploitation in some countries, 
notably Mexico and Equador. About 
60,000 turtles are slaughtered w ^ ^ m- . ^ 
annually in Mexico. n - 3 ' 8: *****<* Gpeen turtte 

The marine Green turtle, Chelonia mydus, found in Myanmar and 
other Asian countries grows up to 400 Ibs in weight and 4-4.5 feet 
long. It lays about 100-200 eggs at a time. Both the eggs and flesh 
are eaten by the local people. 

Crocodiles and alligators are being hunted all over the world for 
their valuable skin. One way to diminish the hunting pressure might be 
to encourage crocodile farming under strict regulations. 

Crocodile farming 

Crocodile farming may be undertaken for commercial exploitation 
of its hide and meat and/or as a way for improving the conservation of 
endangered species and for attracting ecotourism. Cuba has well 
organized crocodile farms that are opened for viewing by tourists. A 
survey should be conducted to determine the status of wild populations 
with respect to their population size, length class distribution, 
location, habitat preference, breeding sites and habitat availability 
prior to the implementation of such farms. An evaluation of potential 
farm sites is also required. Some of the points to be considered, 
depending on the objective of the farming programme, are as follows: 

access to food for adult and young animals 

availability and quality of water; 

access to tourists; 

demand for hides; 

micro-relief and drainage; 

local support and labour supplies. 

Adults are fed waste products from cattle, horse and poultry 
slaughter houses. Small crocodiles are fed "trash fish H (10-12 cm) 
caught by trawlers. There are many successful examples of crocodile 
farming undertaken by private entrepreneurs, and these initiatives have 
tended to reduce the pressure on illegal hunting of adults in the wild. 
However, some of these private farms are known to purchase eggs or 
juveniles from the wild and this should be regulated. 

International trade in crocodile skins and other products is 

largely regulated by the Convention on International Trade in 

Endangered Species of Wild Flora and Fauna (CITIES) and should be 
considered when planning marketing operations. 


3.3.4 Capture fishery 

From an economic point of view, mangroves are often far more 
important for the aquatic production they support than for the wood 
production potential. Kapetsky (1965) estimated that the average yield 
of fish and shellfish in mangrove areas is about 90 kg/ha, with maximum 
yield being up to 225 kg/ha. According to this author, the total 
halieutic production of the world's mangroves would be around 1 million 
tons per year (for an estimated area of 83 000 km' of open water in 
mangroves) , which is slightly more than 1% of estimated total world 
production in all waters per year. 

In the part of the Sundarban Mangroves situated in Bangladesh, an 
average of 9 000 tons of fish and shellfish was caught annually in the 
late seventies and early eighties. With 169 908 Km 2 of waterways, this 
corresponds to approximately 53 kg/ha being caught within the mangrove 
area itself. To this figure should be added the portion of the off- 
shore fisheries of mangrove dependent species. Whereas over 120 
species of fish and shellfish are known to be caught in this area, the 
main portion consists of shrimps (Penaeidae family) and hilsa 
(Clupeidae family) . 

In Africa, Durand and Skubich (1982) report 6 700 tons fish 
landings in 1977 from Ivory Coast lagoons, and Balarin (1984) estimates 
the average annual yield of Benin lagoons for the period 1959-1969 to 
be around 3 700 tons. 

An indication of the strong relationship between fish resources 
caught (including mangrove as well as off-shore fishery) and the extent 
of coastal mangroves is shown in Table 3.7 below. 

Table 3.7: Capture fishery production in relation to mangroves, 1981 


Extent of 


Capture fishery production (x 1.000 tonnes) 1 | 

Overall Non-mangrove 

Mangrove species 1 | 


Mollusca Crustacea F1sh 

West Coast 
East Coast 


433 249 
(100*) (58*) 
216 191 
(100*) (88*) 


71 62 52 
12 12 



649 440 
(100*) (58X) 


71 74 64 

Notes: 1 - Mangrove and non-mangrove resident fishery; 

2 - Includes casual, seasonal migrant and/or mangrove resident species. 
Source: Adapted from Jothy, A. A., 1984 (figures rounded to nearest 1,000), 

Fish statistics do often not separate catches offshore and in 
coastal waters and, apart perhaps from the Sundarbans, accurate 
"Fishing effort data 91 are seldom available for mangrove areas. Such 
data are, however, essential in order to assess the biological and 
economic status of a fishery and to formulate appropriate management 
measures* Data, covering a sufficiently long period, should include 
information on the number of fishing units by various sice classes, the 
catching power per unit and the amount of time spent fishing. 


The state of the fish and shellfish stocks can be determined by 
one of three methods. The first is a simple catch and effort analysis 
based on a surplus production model. The second approach is the so 
called yield per recruit analysis. Basically this approach attempts to 
determine how tie growth and natural mortality of a population interact 
to determine the best size of shrimp to harvest after they have been 
recruited into the fishery. The final method is the analysis of 
species interactions, which is often ignored due to its complexity. 


In Thailand, the main commercial fish species caught in or close 
to mangrove areas include mullets (Liza subviridis) , sea bass (Lates 
calcarifer) , snappers (Lutjanus spp) t tilapia (Tilapia spp) , groupers 
(Epinephelus spp) , sea catfish (Arius spp) , threadf ins (Eleutheronerna 
spp) and snake eel (Ophihctus microcephalus) (Christensen, 1982) . 

The most important fish in West Africa is the "Bonga" (Ethmalosa 
fimbriata) . Other genera of importance inthe same family (Clupeidae) 
are Sardinella and Pellonula. Tilapia is also very important. In East 
Africa, Tilapia and Cyprinus are among the sought after genera, 
followed by mullets, eels and milkfish (Chanos chanos) . 

In Latin America, mullets and snappers are among the most common 
fish caught in and around mangrove areas . 


The term shellfish is used here collectively to describe 
crustaceans (crabs, shrimps) and molluscs (bivalves and gastropods) . 

The main edible crab (Scylla serrata in Asia and East Africa and 
Callinectes latimanus in West Africa) is a highly valued mangrove 
product, and are caught by locally produced traps or by using crab 
hooks to fish the crabs out of their burrows. Other edible crabs that 
are diversely valued depending on countries include some Sesarma, 
Cardisoma and Thalarnita species (SECA/CML, 1987) . 

Shrimps are usually caught with push nets along shallow creeks 
within the mangroves and by off-shore trawlers. In Matang, eleven 
commercial species of shrimps are landed, with the bulk of the catch 
consisting of Jfetapenaeus affinis t M. brevicornis f Parapenaeopsis sp. 
(notably P. hardndcleii and hungerfordi) and Paenaeus 

In a number of countries mysid shrimps are caught for the 
production of shrimp paste, a popular condiment in Southeast Asia. In 
the Sundarbans, numerous fishermen are engaged in catching post larvae 
shrimps to stock the aquaculture operations, as the hatching of shrimps 
is yet in its infancy stage in Bangladesh. Unfortunately, this is a 
very ineffective utilization of the resources as only a few species are 
sought after at this stage and the rest discarded and the mortality of 
the preferred species is high due to storage and transport. In such 
areas care should be taken not to over-exploit the small-sized shrimps 
to the detriment of the more valuable larger adults, which are caught 
by the off-shore trawlers. 


Figure 3.9: Setting traditional fish traps in Matang, Malaysia 

Photo by M.L.Wilkie 

Figure 3.10: Boat load of Blood coddes (Anadan granosa), hfataag 

Photo by M.L.Wilkie 


In Central America the ark shell clams (Anadara sp) are 
undoubtedly the most important economic mollusc resource that thrive in 
the mangrove. In Costa Rica the larger bivalve Anadara grandis 
(chucheca) has practically disappeared due to over -exploit at ion. In 
its place, two smaller ark clams 'pianguas', viz., Anadara 
multicostata and A. tuJberculosa are now being exploited. The latter 
bivalve is commonly found from Lower California to Peru (Keen, 1971) . 
Its importance as a source of protein and an economic renewable 
resource for coastal dwellers makes it the single most important 
exploited species in the mangrove ecosystem of the littoral Pacific. 
Due to the heavy pressure on these species, only bivalves of 45 mm 
diameter and over are permitted to be harvested in Costa Rica. Anadara 
similis, a smaller bivalve is also found and collected in the mangroves 
of tropical America. 

In Malaysia, the blood cockle (Anadara granosa) is a species of 
great commercial importance, that forms the basis of a flourishing 
coastal industry, as well as, being a major protein source. Around the 
Indian Ocean, Gelonia is collected. 

Some wild oysters, as for instance Crassostrea tulipa in West 
Africa, grow naturally on the stiltroots of Rhizophora spp and are 
locally exploited. It should be noted however, that oysters can be 
dangerous to eat during outbreaks of the so-called red tide. 

For information on the biology and culture of Anadara sp. and 
tropical oysters, refer to Broom (1985) and Angell (1986) respectively. 

Other groups of bivalves harvested and used as food are the jack- 
knife clams and mussels. The clam, Polimesoda in flat a, is closely 
associated with certain crabs (Pinnotheres sp) . Modiolus capax 
(Mytillidae) has a triangular shell and bundles of filaments that 
enable them to attach on to mangrove roots. 

Various species of edible snails are found in the mangroves, and 
some of these are locally commercialised. Terebralia and Telescopium 
are usually eaten by cracking their shells, while the insides of the 
smaller Ceritbideae are taken out with a toothpick or simply sucked out 
when the closed end is broken. Species of JVerita and Salinata are also 
gathered as food. (Christensen, 1982) . 

3.3.5 Mariculture 

Traditionally, mariculture, involving the use of a system of man- 
made ponds in rearing specific marine or brackish- water animals, has 
been practised in Indonesia for hundreds of years. Ponds "tambaks" 
were constructed to rear milk-fish (Chanos chanos) . Along the mangrove 
waterways, creeks and estuarine waters, a rich tradition of artisanal 
mariculture has evolved and fish constitute an important part of the 
peoples' protein supply. 

Unlike many cultures of .Asia, the Pre-Columbian cultures of 
America had no significant tradition of aquaculture, nor did the 
Europea'riS* - who came in the fifteen century. Consequently, there are few 
mariculture traditions. The preferred protein is meat rather than 
fish, where only a few species are consumed. 


Several mariculture practices are used in the mangroves, and 
these may broadly be classified into those methods which make use of 
the natural fertility of the estuarine aquatic system without 
destroying the vegetation (open-water estuarine culture) and those that 
are practised on the land (pond culture) . 

Open-water estuarine mariculture 

Three main types of open-water estuarine mariculture can be 
distinguised viz. Bottom culture, where no enclosures are used, Cage 
culture and Raft and cultch culture. 

Bottom culture 

The genus Anadara has wide mariculture potential and the Blood 
Cockle (Anadara sp.) is a good example of bottom culture* The 
substratum and the exposure period at low tide appear to be the more 
important factors limiting the distribution of cockles. Anadara 
granosa is a major protein source, that grows naturally on mangrove 
mudflats in West Malaysia, Thailand, Kampuchea and South Vietnam. 

Cockle farmers collect the seeds (measuring 6-12 mm) from 
spatfall areas on the higher mudflats at low tide and sow them in the 
cockle beds in the lower areas up to 4.5-5.5 thousand litres/ha. 
(Hamilton, L.S. and Snedaker, S.C., 1984). Predators such as starfish 
and Natica are removed at low tide. Sown seeds mature after 8-12 
months and are harvested when they are about 3 cm in diameter. The 
yields in the better areas are about 20.7 and 24 t/ha/yr in Malaysia 
and Thailand respectively (Sribhidhad, 1973) . Malaysia is the largest 
producer and exporter of cockles among the tropical countries, with 4 
700 ha of cockle beds producing about 65 000 t annually (equalling an 
average 13.8 t /ha/year ), valued at over US$ 12 million. 

Anadara tuJbercuiosa found in tropical America is another species 
which offers excellent prospects for commercial development (Ellis, 
1968, Hagberg, 1968) . 

Bivalves being filter feeders are very sensitive to water quality 
and contaminants. In 1970, for example, waste discharge from sugar 
refineries into the Mae Klong river in Thailand practically destroyed 
the cockle farming industry in the estuary. Reportedly, bivalves in 
the Sierpe mangroves in Costa Rica were reduced drastically due to 
copper-based pollutants discharged by the banana plantations. 

Seaweed culture 

China, Hongkong, Vietnam, The Philippines, Taiwan, Japan and 
Korea are the Asian countries that consume large quantities of seaweeds 
as food, and also for medicine and cosmetic purposes. 

In Thailand seaweed farming based on Gracilaria is carried out in 
shallow coastal waters along mangrove shore lines, where the bottoms 
comprise silty sand. Large quantities are exported to Japan. In West 
Malaysia, the best seaweed farming sites are located inshore of the 
shrimping grounds, in the lower intertidal zone. 


Seaweed farming is also practised in the Philippines with good 
financial results and similar potentials exist in Vietnam and parts of 
Myanmar. The synergic relationship between seaweed and cockle farming 
has not been established, but given the generally known ameliorative 
effect of sea-grass meadows on aquatic productivity, these two 
activites are likely to be mutually beneficial (FAO, 1977) . The 
biggest constraint now and for the future is the growing pollution of 
coastal waters. 

Cage culture 

In the sheltered estuaries and canals that are rich in organic 
detritus, the rearing of fish can be carried out in cages and 
enclosures. The use of cages made of synthetic net or bamboo screens 
measuring from 0.25 ha to 5 ha each have been successful in the culture 
of milk- fish (Chanos chanos) in the Philippines, producing as much as 
4 t/ha/year. Some supplementary feeding was provided. (Delmendo and 
Gedney, 1974) . 

Floating net cages are suitable for species which can tolerate 
crowding, convert feeds efficiently, are easily available as fry, are 
highly priced and in good demand (Christensen, 1982) . The two most 
cultured fish are the sea bass (La tea calcarifer) and grouper 
(Epinephelus tauvina) grown in cages (4-5 by 5-6 m and about 2.5 m 
deep) . The stocking rate of fry of these two species range from 350- 
500 per cage. In Malaysia, a yield of about 75-125 kg can be obtained 
from each cage after 10-12 months (Chan and Salleh, 1987) . 

Raft and cultch culture 

Oysters and mussels have been grown on nylon ropes suspended from 
floating rafts, yielding about 180 t/ha/yr of mussels. 

In Lower Allen Town close to the capital Freetown in Sierra 
Leone, women used to gather the wild oysters (Crassostrea tulipa) 
during low tide by cutting the mangrove roots. Due to the destructive 
method used and overcutting for firewood, the coastal mangroves have 
been transformed into low shrubs and even destroyed, leading inter alia 
to a decline in the amount of oysters collected. However, by stringing 
oyster shells together with a nylon string, and hanging these on bamboo 
racks by the tidal creeks, oyster spat can be collected and mature 
oysters harvested without destroying the mangrove vegetation after 
about 12 months. This method has been applied with some success in the 
Lower Allen Town area (Chong, 1989) . 

In Thailand, concrete cylinders, 15 cm in diameter and 40 cm 
high, are mounted on short Phoenix palm posts at a density of about 1 
post/m* . The spats of Crassostrea commercial is attach to the 
cylindrical surface and can be harvested after 8-18 months. 
Christensen (1982) reported that about 17 t of meat could be produced 
per hectare per year with this method. In Panama and the Caribbean sea 
coast of Costa Rica, the mangrove oyster C. rhizpphorae has been the 
subject of research. The seed could be collected from natural 
reproduction and planted in other areas for growth to market sizes. 
Culture methods have been well established in Puerto Rico, Cuba and in 
Bocas del Toro, Panama. 


Figure 3.11: Cage culture, Matang, Malaysia 

Photo by M.L.Wilkie 

Figure 3.12: Oyster culture along a mangrove creek, Stara Leone 
Photo by P.W.Chong 

Pond culture 

-scale ond clture 

Artificial fish ponds, comprise about 95% of the aquatic culture 
in the Philippines/ where about 3,700 ha of mangroves were destroyed 
annually during 1952-1981. (Utaali, 1985) . Ponds about 0.1 - 1.0 ha are 
constructed within the mangroves to take advantage of the availability 
of fry, natural fertility of the mangroves and tidal flushing. 

Another example of this is the % tambak' fish-ponds in Java where 
mangroves and other useful trees are often grown on the dikes between 
the' ponds. (Sukardjo, 1978). Sometimes, Nipa is also grown beside the 
ponds to provide shade. The ponds are mostly used for the culture of 
shrimps and milkfish. 

Milkfish (Chanos chanos) fry is common in mangrove environment 
and coastal waters in Southeast Asia. It is generally caught after the 
post-larval stage and before the fingerling or late fry stage. 

Large-scale pond culture 

Mangrove areas in many developing countries are increasingly 
being converted into large a qua cultural ponds used mainly for rearing 
shrimps rather than fish due to high export demand and shrimp prices. 
This occurs particularly in areas where the coastal waters are rich in 
nutrients, stocked with wild post-larvae and juvenile shrimps of 
commercial species, and the tidal range is favourably high (about 3 m) . 
Shrimp-ponds are constructed in the mangroves because the sheltered and 
shallow estuarine areas are the natural habitat of a variety of 
commercial wild shrimp, providing gravid females and abundant 
post larvae and juveniles. 

The postlarvae of the Indo-Pacific 
species Penaeus monodon, P. indicus and P. 
mergruiensis, and the eastern Pacific species 
P. stylirotris, P. vannamei, and P. 
occidental is are normally found in the tidal 
creeks and do not migrate into the 
mangroves. Wild postlarval and juvenile 
stocks are declining in numbers due to the " 3.13: Shrimp 
continuing degradation and destruction of 
their habitat. 

The technology for producing postlarvae shrimps in hatcheries 
exist, but it is expensive, complex and not available to small 
operators. Consequently an adequate supply of fry is becoming a 
problem for small operators who cannot afford to buy stock from the 
hatcheries. The dependence on unreliable wild fry supply is the 
weakest link in the production chain for small shrimp operators. 

Compounding this problem is the cyclical occurrence of the "el 
nino* phenomenon, which causes the coastal seawater to warm up. This 
has caused hardship to shrimp farmers in Central America, due to 
flooding and destruction of their ponds, salinity changes in coastal 
waters and steep decline in wild fry supply. 


Figure 3.14: Clearing of mangrove area for shrimp ponds, Malaysia. 

Photo by M.L.Wilkie 

Figure 3.15: Monitoring the shrimp production, Malaysia. 

Photo by M.L.Wilkie 


In Panama, where mar i culture practices are quite sophisticated, 
production is constrained by the high salinity (47-50*) experienced 
during the dry season and the shortage of wild fry. An account of the 
major problems! as reported by Cintron, (1985), is summarized below: 

Availability of seed: Wild seed stocks are unpredictable and low during the dry 

months. P. vannamei is plentiful during the wet period, while P. stylirostris is 

more during the dry season. 

Feeds. Supplementary feed pellets constitute 20-40% of the annual production 


High cost of oostlarvae: Laboratory seed is 250% costlier than wild seed. 

Postlarve stocking accounts for 1 6*20% of production costs. 

Climate: Salinity becomes excessive (>50%o) in dry months (January-April), 

and bunds are often damaged by heavy rains during the wet season. 

Diseases: These are becoming more common as the intensity of production 


Predators: Aquatic predators and birds reduce yields. 

Contamination: Coastal and estuarine waters are becoming more polluted. 

Theft: Security costs must be considered. 

In the Philippines/ Rabanal (1977) recommended lands which are 
flooded during ordinary high tides and drained at low tides as the most 
favourable for pond construction (Figure 3.16 on the following page) . 

3.3.6 Salt production 

Solar salt production is a traditional and important industry in 
many coastal dry and semi-dry regions. As a basic commodity salt is 
required in the human diet and in some agricultural and industrial 
applications. It is used also for preserving fish, beef, fruits and 
vegetables. In 1980 about 25% of the world's salt production of some 
175.5 million tonnes was produced using solar energy. 

Seawater is guided into and trapped in bunded ponds constructed 
on higher ground during spring tides. Upon evaporation the salinity in 
the evaporation ponds increases until salt crystals precipitate from 
the concentrated brine. Few plants can survive under such hypersaline 
conditions. Development of salt flats depends on the following factors: 

supply of seawater with high salinity; 

a pronounced hot and dry season when potential evapotranspiration exceeds 

precipitation (PET > P); 

flat coastal land; 

restricted surface/underground freshwater inflows that can dilute or leach the 

accumulated salt/brine; 

prevailing drying wind to accelerate evaporation. 





(in dm) 























Fig. 3.16: Lands suitable for pond construction in relation to tidal 

in the Philippines. (Rabanal, 1977). 

Solar salt can be produced more efficiently in semi -arid 
climates, or in regions having a pronounced dry and hot season, when 
the potential evapo- transpiration exceeds precipitation. Equatorial 
climates are not favourable as evidenced by the aborted Sungei Merbok 
scheme in West Malaysia. In the arid zones, salt can be produced 
throughout the year. Solar salt production is generally situated on 
the seaward side, where mangroves are climatically restricted and 
"salt-flats" (known as alJbinas in Panama, salitrales in Cuba and tannes 
in parts of West Africa) develop naturally. 

The following description of the salt -making process is extracted 
mainly from Cintron's report on Panama (1985). 

Modern saltworks comprise several evaporation ponds where the 
brine is evaporated to three broad salinity levels* In the 
precon centra tors, sea water of 30-40fc is evaporated to about 80k, at 
which point ferric oxide and some calcium carbonate precipitate out. 
The brine is drawn into concentrators, where its salinity is 
concentrated to about I80fc, removing the rest of the calcium carbonate, 
borates and calcium sulphates; and "nodrizas" where, with salinities 
greater than I90lb, the remaining calcium sulphate is removed as 


Finally, the concentrated brine, which may be kept in holding 
tanks, is fed into "destajos" or crystallizing ponds, where sodium 
chloride precipitates are collected and further dried under the sun. 
All the salt is sold to the government. 

The profits of the saltworks may be augmented by raising brine 
shrimp (Artemis sp.) as a side operation. Brine shrimp occurs 
naturally in waters of 80-260fc salinity, but develops best between 
140- 180k. Artemia is sold to shrimp farmers and the cysts are in great 
demand (OS$ 70/kg.) . During the wetter months, ordinary shrimp may be 

In Punt arenas, Costa Rica, an effort was made to utilize existing 
salt ponds "salinas" for the production of mullets or liza (Mugil 
curema) a low value product for the regional market and shrimp, a high 
value product, for local consumption and for export. 

In Sierra Leone, which has an equatorial and very humid climate 
salt making through solar evaporation is not possible. However, great 
quantities of salt are produced in the mangrove areas in the dry season 
with a slightly different technique: The top layer of the mangrove 
soil is scraped off and loaded into a V-shaped vessel lined with straw 
and mud and leached with sea water. The resulting brine solution is 
boiled in big pans until the water evaporates and the salt is left as 
a residue (Figures 3.17 and 3.18) . This method, if practised on a big 
scale, can be very destructive as mangrove wood is used for the boiling 
process. However, in most areas in Sierra Leone it is undertaken on a 
small scale in the drier, less productive parts of the mangroves and 
utilizing Avicennia wood, which regenerates through coppicing. 

3.3.7 Agriculture on mangrove soils 

Generally mangrove soils are marginal for long term agriculture 
due to the chemical nature of the soil, salinity, and shrinkage and 
subsidence when the soil is tilled. During the dry months, shortage of 
potable water causes hardships and sanitation problems. 

Farmers in tidal areas experience a number of problems related to 
their specific environment and soil conditions. In the first place 
because the land is flooded during high tide, a system to keep out the 
saline water must be devised. This takes the form of low bunds, dikes 
and/or drainage canals. Dikes would curb the flood hazard in shallow 
levee of rivers but the construction is expensive. Close to the shore 
saline or brackish water creeps in from the sea (salt intrusion) . 
Where mud bunds are constructed, these have to be continually 
maintained especially in areas where the mud- lobster (Thalassina) is 
common. Yields are low when the soil acidity is high. 

Due to the high incidence of potential acid sulphate and acid 
sulphate conditions in the mangrove mud, an adequate supply of water is 
not only necessary but water control to maintain the water-table above 
the sulphitic layer is a precondition to the successful reclamation of 
mangrove land. The field may be rain-fed or irrigated. This will keep 
the soil moist for most of the time, but this also limits the range of 
crops that can be planted. 


Figure 3.17: Leaching of mangrove soil with salt water, Sierra Leone 

Photo by courtesy of M.P.Wilkie 

Figure 3. 18: BoiHnf of the brine solution to Mf pro, Stern Leone 

Photo by courtesy of M.P.Wilkie 


Large-scale reclamation of mangrove lands for agriculture 
requires very careful planning, including a realistic economic analysis 
(See o*e 3.2 below) and an analysis of its impacts on other resources. 
Additionally, where the land is close to the sea and prone to cyclonic 
storms, the toll in terms of human lives and destruction of property 
and crops could be very high. 

to determine the econoaiic viability of reclamation of 

' ' ' 

discount rate df 5 % and 

, , , ' 

With regard to reclamation for sugar cane production without 
irrigation aod with a Bttell shritnp pond coropOTient the Net Present 

1H$#* was ;i^*fciv ^ ^ttoti talcing into account the net 
benefits forgone for forestry and fishery. 

Concerning the reclantation for rice cultivation with irrigation 
using salt tolsramt HYV rice, the NPV was also negative when 
undertaken, on mangrove soils - again without taking the net 

benefits f:0*0o**0 into account. 

The main reasons were the problems incurred due to the formation 
.iis^itt^ roll* including low yields and the special farming 
practises iweded and the costs of bund wall maintenance. 
The SPV estimates for the above were still negative under the 
assumption of accelerated desalination process and using higher 
social discount rates and lodger planning horizons* Only at one 
to zero discount rate and 100 year planning horizon with no further 
capital investment did the rice project give a positve NPV. The 
PV for sugarcane was however still negative. 

It was thus concluded: 'if the effects of ecological 
characteristics of mangrove soils, which determine the time taJten 
for freshly reclaimed soils to become productive, were to be 
incorporated in the benefit -coat analysis, then even wit 23 
rice cultivation is act economically 

Box 3*2: An economic analysis of mangrove reclamation for agriculture in F\ji. 

Paddv cultivation 

Rice cultivation became a major food crop in mangrove areas in 

Guinea and Sierra Leone aroung 1855 and much later in Madagascar 

(1935) . In Myanmar (1852) , the colonial administration, who considered 

the luxuriant Kanazo (Heritiera femes) forests to be "wastelands", 

transformed the Ayeyarwady delta mangroves into paddy fields. 

Traditionally, the Buginese and Banjarese farmers in Kalimantan 
and Sumatra have planted rice along the tidal coastal swamps for over 
seventy years* Rice yields of up to 3 t/ha are produced on highly 
#yritic soils, due to careful water control and tidal flushing of the 
top soil. The farmers use a system of closely spaced parallel shallow 
drains that effectively removes any acidity developed in the surface 
layer while keeping the subsurface saturated with water. 


Apart from toxic levels of iron and aluminium ions, the level of 
available phosphorous is very low. Copper, Zinc, and Manganese contents 
are also low in many coastal soils. In any case, the adequate 
maintenance of canals and waterworks is a precondition for sustained 
rice production. 

The four main types of rice cultivation systems are summarised in 
Table 3.8 below. 

Table 3.8: Rice cultivation systems 




Flooded rice cultivation 

1. Tidal rTcel Freshwater discharge during the rainy season for a period 

of more than 100 days 1s necessary. Flooding during high 
tide. Mostly executed 1n upper estuarlne zone; 

2. seasonally flooded rice: Sufficient discharge (flood -height) 1s 

necessary to maintain fresh water Influx 
and levels during the growing season. 
Rain-fed rice cultivation 

"Rainwater polders": DurTno the rainy season high water tables are maintained 
within small basins. Rice 1s often cultivated on small ridges. At 
least 1,000 mm of rainfall during 4 months 1s needed. During the 
dry season the fields are kept moist with brackish water to avoid 

Control 1 

high water-table agriculture 
n humid climates wit 

table aqr 

sites with no pronounced dry season high water-tables 
are maintained with minimal drainage. In the tropics often used for 
palm-oil cultivation. Intrusion of salt-water is controlled by 
sluice gates; 

Total reclamation: 

:omplete control of drainage and Irrigation. Soils have to be fully 

Source: Adapted from Dent. 1986. 

The International Rice Research Institute (IRRI) has classified 
mangrove lands according to their suitability for paddy cultivations as 
follows : 

Most suitable: 
Moderately suitable: 

soil salinity 1s low, rainfall 1s high and well 


soil salinity due to evaporation and rising of 

the water table is not very high; 

high salinity, acid sulphate soils: soil exposure 

and dryness in the dry season and deep flooding 

1n the wet. 

Apart from irrigated rice, shallow rooted crops such as 
vegetables, or oil palm and coconut are the only crops suitable on 
these reclaimed areas, especially in the initial stages of reclamation. 
On mangrove swamps which have been reclaimed for a longer period, mango 
and other fruit trees have been established successfully. Rubber does 
not yield well on these marine clay soils, due to the high amount of 
exchangeable calcium and magnesium. 


Coconut has often been the first crop planted on reclaimed 
mangrove swamps, due to its salt tolerance. Yields are however 
affected by pH changes brought about by drainage especially in the 
potentially acid sulphate and acid sulphate soils. Experimental trials 
showed that a drop in pH caused by drainage resulted in yields 
plummeting from 5 154 to 2 818 nuts/ha/yr (Zahari, 1983) 

Sometimes cocoa and bananas have been successfully intercropped 
with coconut on acid sulphate soils, assisted by the application of 
limestone powder around the palm clumps. 

Oil palm 

In Malaysia potential oil palm yields (about 55 t of ffb/ha/yr) 
from reclaimed mangrove soils are higher than those planted on upland 
soils (average 20 tonnes ffb/ha/yr) . (Poon, 1983) . Proper water 
management is the key to successful cropping and the present drainage 
control system used in Malaysia, which aims to maintain the water-table 
at about 60 cm from the surface of the soil, has overcome the problem 
of the formation of acid sulphate soils. (Poon, 1983). 

3.4.1 Coastal protection 

Coastal protection is often referred to as either "shoreline 11 or 
"coastline 11 protection. The measures taken in the form of seawall 
barriers and other ant i -erosion structures are collectively referred to 
as seawall defence in this document. The coastline is defined 
specifically as the high spring tide shoreline. 

Many wetland and mangrove areas are destroyed due to the 
construction of protective structures. These are frequently located at 
or seaward of the estuarine low water mark. Large areas of productive 
wetland and tidal flats may be sealed off from marine influences and, 
likewise, the seaward areas of the estuary or lagoon are deprived of 
freshwater inputs. Highly productive wetland/tidal swamps areas may be 
lost. Refer to Box 3.3 on the following page. 

MitJ9atorv measures 

Prevention is better than cure. This particularly applies to 
coastal protection, where the remedial costs can be phenomenally high. 
Structures should be placed behind the line of annual flooding which 
marks the landward limits of coastal wetland. Walls should be 
constructed from rubble { "rip-rap") to allow the movement of water- 
borne nutrients from inland areas to the estuary. Rip rap has the added 
advantage of reducing wave scour and providing new niches for estuarine 
organisms . 


Guyana's dttamwi cnanfataf a***) defence 

Guyana cannot afford to ignore the sea. Not only Is sugartte heart of it* economy, 
but 90 per cent of It r s 750,000 people live atong the narrowly stretched coastal wme 
that constitutes less than 3 percent of it's total land area. This narrow strip /600km 
tang and never more than 16 km wide, carries the most fertite soils; the rest of the 
country has marginal or no agricultural potential, This coastal strip that accounts for 
over 70% of the country's GDP unfortunately lies below the Wgh tide mark, making 
it particularly vulnerable to flooding, seas erosion and 

Coastal protection, however, is expensive consuming more than 30% of the 
country's capital expenditure up to the mid- 1970s. However, (Hie to economic 
recession, most seawall defence are neglected, requiring complete replacement in 
many places, 

As if keeping out the sea water problem is not enough, a rise in sea level of even 0.6 
m due to global warming will flood most inhabited areas. A sea level rise of 1 .5 m 
could potentially destroy an income of $107 m from the sugar industry, $46 million 
from rice, $84 m from other crops and $11 m from livestock; putting about $800 m 
of economic activity at risk. 

To replace even the worst affected 130 km of the coastal defence would require 
$260 m, A typical type of physical protection required would cost about $2 ODD a 
metre to build. 

However, experience gained by an ODA project on sea wall defence in the vicinity 
of the Essequibo River between 1979 and 19&4 indicates that the "best coastal 
and 9 sfMH earthen dam betmJ that". The foreshore stretches out to sea, 
sometimes for 5 km or more, and mangroves at one time protected large part of the 
coast. The penciHike aerial roots of the black mangroves Avfeyfofrffl. dissipate 
much of the tidal and wave energy. The water is retatively catm, gerrtty rising and 
falling with the tides. The early Dutch settlers took advantage of the mangroves, and 
only needed simple earthen darns, about 1.5 m high and 4 m wide at the crest, to 
keep out the gentle waves from flooding the adjacent agricultural land, 

The mangrove protective belt has deteriorated due to: Jiisik, the tack of cooking fuel 
has forced many people to cut mangroves and agfiflffiflfr the natural cycle of erosion 
and accretion resulting from the body of mud that emanates from the Amazon, 
coupled with coastal currents. The Guyanese refer to it as sling mud, The very fine 
mud particles, always remains in suspension and tends to clog up the "breathing 
pores' 1 of the mangrove pneumatopfeores, except for the hardier 

Box 3.3: Coastal protection in Guyana 


It may be possible to circumvent the need for any kind of 
structure by planting tidal vegetation as a flood prevention and 
erosion control measure. These are much cheaper and almost maintenance- 
free when fully established. 

Avicennias are choice candidates because of their hardiness, 
coppicing ability and as they are able to withstand high salinity. 
However, potted seedlings may have to be used and appropriate 
protection provided. 

Low mangrove shrub vegetation bordering estuaries and low energy 
coastlines can be established to form a protective barrier against sea 
attack. They decrease erosion along unconsolidated coastlines by 
breaking the force of waves and dissipating wave energy. In estuarine 
locations, mangroves perform the additional role of trapping silt and 
gradually raising the level of the shore. Dense thickets of mangroves 
present an effective barrier to storms along low- lying tropical and 
sub-tropical coasts and may be planted specifically for this purpose. 
Box 3.4 on the following page describes such an example. 

3.4.2 Recreation and Ecotourism 

Tourism accounts for one-third of the trade in goods and services 
of developing countries and the World Tourism Organization (WTO) 
projects that it will become the world's largest industry by the year 
2000 (WTO, 1989) . There were 390 million international tourists in 
1988, who created 74 million jobs and produced $195 billion in local 
and foreign receipts. Adventure travel, which includes ecotourism, 
commanded almost 10% of the market in 1989 and is increasing at the 
rate of 30% a year. (Kail en, 1990) . 

Ecotourism potential can only be realized if the resource on 
which it is based is well protected. In turn, it can empower local 
communities, give them a sense of pride in their natural resources and 
heritage and control over their communities' development. It can 
educate travellers about the importance of the ecosystems they visit 
and actively involve them in conservation efforts. In sum, it has the 
potential to motivate rural population, maximize economic benefits and 
minimize environmental costs. 

Activities that may be promoted in mangrove areas are nature 
trails, bird watching, nature photography, crocodile -farms, fishing, 
river rafting/canoeing/kayaking, and botanical studies. 

Ecotourism is an excellent economic option, with positive and 
wide-ranging social, political, and environmental benefits. However, 
it should not be viewed solely as an avenue for short-term financial 

The requirements for making ecotourism profitable and beneficial 
over the long term are, firstly, to train tour promoters and 
operators; secondly, to select sites which provide appropriate 
recreational and educational opportunities for visitors; and lastly, to 
ensure that activities are environmentally compatible. Low impact 
tourism is less intrusive than resort tourism, and therefore, more 
suitable for protected areas which cannot sustained any direct use. 


- '. . Bantingofnungrovcs tor protection 

Each year several typhoons and strong storms hit the coast of Vietnam, causing 

worst affected ara is SttMUtad iin the Central Region of the country, from than Hdi 
province in the North to Quang Nam Da Nang province in the South. 

These typhoons are accompagnied by increases in see level of 0.5-2,5 m and by 
waves with amplitudes of two metres. The existing dykes were unable to contain 
or withstand such forces and were breach every one or two years. In order to 
alleviate the problem, assistance was sought from the World Food Programme to 
upgrade the dykes through a Food for Work project. 

A forestry component was included in the above project with the following objective: 

To protect the rehabilitated and upgraded dyke systems to 7 provinces 
against erosion by wind, wavas and watar currants through the planting of \ 
trees in front of dykes. 

This objective reflected past observations on the positive effects of a tree cover on 
the seaward side of the dykes. The occurrence of such natural stand* in front of 
dykes have, according to local farmers, in many instances resulted in only minimal 
damage to dykes by typhoons, whereas neighbouring dyke sections without this 
protective barrier suffered severe damage. 

It should be mentioned here, that although the dykes are referred to as coastal dykes 
or sea dykes, they are in fact not located along the coast itself, but along rivers, 
estuaries and lagoons which are uderthe influence of tidal water from the sea. There 
is thus normally no direct breaking of waves in front of the dykes. 

A total of 454 km of dykes are to be rehabilitated and upgraded under the project and 
along 195 km of these trees are to planted in front of the dykes for protection 
purposes. The remaining areas are either unsuitable for the establishment of 
mangrove plantations or do already possess nature! mangrove stands. 

The species to be planted are Rhizophora sp., KandeJia candet and Nypa fruticarts 
with a few hectares of Filao (Casuartna sp.) on higher elevations and sandy soil. All 
of the above species can be found within the protect area. 

The width of the areas planted depends on the soil type, the topography and the tidal 
regime. Whereas a width of 50-1 00 m is desirable, the average wktth wiH be 30-50 
m due to the above constraints in particular the steepness of the shores. The spacing 
adopted is very dense with the major part of the plantations to be established it 1x1 
m equalling 10 000 trees/ha. 

A total of 1 010 hectares are to be planted during this project and a mechanism for 
maintenance and protection of the plantations by the local communities has been 

Box 3.4: Protection of coastal dykn by mangroves ta Vietnam 


Some recommendations for low- impact activities in mangrove areas 
are summarized in Box 3.5 below: 

ntfvcy to dotanubifr ; tto idaaottfti tod : y0ts!tjy ite 
&e penmttibte carrying cap*** for hamam, boat*, wildlife, damettte animals 
should be ckrtenmncd ai^ enftmcd; 
corfrol poaching jus illegal hunting of wMifc; 
observe itrict hypeoe to avoid the spread of water-borne disease*; 
ttgjtf^ for afl tourism <tevelopmei* prcrjectt 

that teve the potential to degrade muni and cultural retoufce*; 
integrate tourism development ptanmng with other agencies; 
incorporate visitor management into the area management plan; 
continually monitor the aites, identify impacts, and take measures to eliminate 
eoviroittneQbtl degradation; 

include a strong eovironmentol education component that provides guidelines for 
"tow-impact tourism", stimulates in ecosystem awareaeas, and provides for direct 
partfc^ation in conravation efforts ind tourism programme*; 
maintain close links between the local communities concerned tad mangrove 
resource managers to well protect and manage the tourism resource; 
information- and data-gathering efforts related to tourists and tourism should be 
improved tod stindantized. 

Box 3.5: Guidelines on planning eootourlsm In mangroves 

pcotouriam and its role in sustainable development 

People must be provided with simple and viable alternatives to 
destroying their natural resources. Ecotourism has the financial 
potential to provide a viable economic alternative to the exploitation 
of the environment. If properly organized, ecotourism can be 
sustainable business at the national and local levels. It creates 
local employment and income to local communities, as well as foreign 
exchange to national governments, while conserving the natural resource 
base in a productive manner. 


3.5.1 Integrated coastal area planning 

The mangrove swamp is closely linked to terrestrial land use 
practices. In particular, changes in water-flow regimes affect' the 
mangroves, and the overdrawing of groundwater or excessive removal of 
mangrove vegetation may increase the danger of aquifer salinization and 

Consequently, the coastal zone should be considered as an 
integral component of overall regional land use planning and 
development so that appropriate land use policies and action programmes 
may be formulated. Priority should be given not only towards the 
rehabilitation of degraded coastal lands but also the rational use of 
land on a sustainable basis, including the planned development of 
sustainable forest/marine products. 


Through in -situ rural socioeconomic development programmes geared 
towards optimizing the use of available resources, the coastal zone 
will be rendered more productive and environmentally stable. 

Many of the uses and services of mangroves are compatible such as 
for instance forestry, honey collection, coastal protection and small 
scale capture fishery. Others are less so and a zonation of the area 
according to primary land use objectives might be necessary. This 
underscores the need for a holistic approach within the framework of 
integrated coastal area management planning. 

3.5.2 Land use issues and conservation 

The main management land use options in mangrove areas are: 

(a) Conservation; 

(b) Sustainable utilization, and 

(c) Conversion to other non-wood uses. 

(a) Conservation is defined as "the management of human use of the 
biosphere (i.e. all living things) so that it may yield the greatest 
sustainable benefit to present generations while maintaining its 
potential to meet the needs and aspirations of future generations" 
(ICUN, 1980) . 

In short, conservation means the maintenance of all living 
resources and management in the use of such resources. Once 
conservation is recognized as maintenance of the means of development, 
then, with proper planning, it becomes possible to integrate the two 
processes/ viz., maintenance and utilization and hence to make 
development sustainable. Integration, however, depends first on 
recognizing the roles that ecosystems play in human economics. 

Protected Areas: Ecologically unique and fragile ecosystems should be 
preserved as "protected areas", not only to afford protection to 
endangered wildlife and flora but more importantly to maintain 

(b) Sustainable utilization; Single use options should be avoided 
because they sub-optimise the multiple-use potential of mangrove 
ecosystems. Wood removal for instance does not conflict with capture 
fishery and open- water mariculture, if the character of the forest is 
preserved and timely regeneration measures are undertaken. 

(c) Conversion: Outright conversion to non-wood uses forecloses many 
of the biological and financial advantages that a natural ecosystem can 
offer. In some cases the changes become irreversible due to 
biophysical factors or simply because the restoration costs are too 

To minimize adverse impacts on the mangroves, the following 
suggestions are made: 


(1) The mangrove swamp is intertidal and becomes progressively higher 
in the tidal zone as it develops. It is suggested that only the 
driest mangrove swampland be developed for non-wood purposes, 
such as large-scale pond mariculture or agriculture, as such 
areas are marginal for wood production. 

(2) It is recommended that potentially acid and sulphate soils should 
be avoided for paddy cultivation. If possible, consider only 
soils that do not require any reclamation. Excessive clearing of 
the natural vegetation should be avoided. Fresh water must be 
assured in adequate amounts and proven salt -tolerant varieties of 
rice should be used. 

(3) Where land reclamation for industrial development is 
contemplated, the negative ecological effects and economic costs 
associated with the loss of natural coastal protection provided 
by the mangroves and the possible decrease in fishery revenues 
and other benefits should be evaluated. 

(4) Before conversion is undertaken an environmental 
assessment should be undertaken. 


Figure 3.19 below illustrate what should be avoided. Examples of 
integration of uses are described and illustrated in Cast Study 6, 
Box 3.6 and Figure 3.20. 

Figure 3.19: Effect of road on mangroves, Cuba 
Photo by P.W.dxmg 


In the Ngoc Heia district in Southern Vietnam, mangroves are seen as habitats for food, shelter 
aad as land to carve out aa economic tawe for the many settles 

<3&<*tata forests ftteog waterways have been converted ttuhrfcnp tetotflfc, 

production varies from 150-350 kg/ha. Declining yields in more than 45% of (he estabtohed 
pond* doe to pond acidity, poor management, reduced posOarvae supply tort impoverished 
foreat-cum-aquatic habitats are forcing some fanners to expand the size of thek bofcttii|s ar 
create new ponds. A seven fold increase in pood area was registered during 1999-4& Tbe 
auraber of households within the area rose ftom 53 to W during the saioe period, tftricts n 
ecological balance is struck between ihrimp-farming, fishery and forestry developmrat ^*Jmi a 
sustainable context, there will be no economic nor eavkonmcctal ftmire for those who oB the 
land now and for their children, 

pian was pft(Mtffed> 
afeo advocates the 

To optimize the mukiplc use potential of the mangroves, a management 
which considers not only tree resources and shrimp taming tat 
tstabtishmtju of growth centres and forest villages as one of the means of marahafling available 
human resources into ecologically viable sites, so Oat social services and infrastructures can be 
provided in a cost effective manner At the private farm level, a simple shrimp-tree fanning 
model was proposed based cm four hectares ofwoodlot to about oae hectare ofgrownttit ponds. 
(Chong, 1988; Karim, 1988). Private woodlocs, managed 00 9-12 year rotation*, tie situated 
behind the fish/shrimp ponds. Access and irrigation canals (1.2m deep) at intervals of 250-280 
m for extracting forest produce and also to guide tidal flows into the panda are provided. Each 
household is allotted 2 ha of hod for shrimp ponds and house and 8 ha ofwoodlot. A2ft4Qm 
protective mangrove belt bordering the waterways is twined. Refer to Figure 3 M below. 

Box 3.6: Integration of uses in mangroves in Vietnam 

, . Rotation 12 years \ 

i fi i. \ .5 :. . t\ 


tot Mui Fovut Enterprise 
m*ngrov*s - notation 30 ytars 

k . L 



Fig. 3 JO: A 

fanning mo<W (UNDP/FAO:VlE/82/e2) 



The socio-economic value of mangroves stems from a variety of sources 
some of which are difficult to value as they either are services and goods 
which are non-marketed or they occur off-site, that is they are economic 
externalities . 

Three examples at attempts to put a value on mangroves are described 
below and in Box 3.7 on the following page. 

ESCAP (1987) estimates that the probable direct employment offered by 
the Sundarbans in Bangladesh is likely to be in the range of 500 000 * 600 000 
people for at least half of the year, added to which the direct industrial 
employment generated through the exploitation of the forest resources alone 
equals around 10 000 jobs. The Bangladesh Forestry Department obtained 
revenues from the Sundarbans equalling Tk 140 mill, in 1982/83, but this is 
a severe under estimate of the value of the area, as some of the royalties 
collected were exceedingly low. For Sundri (Heritiera fames) fuelwood for 
instance, the market rate was nearly 40 times the royalty rate and for shrimps 
the minimum market rate to royalty rate ratio at the time was 136:1. 

Tang, Haron and Cheah (1980) estimated the market value of a 30 year old 
mangrove stand in Peninsular Malaysia, clearfelled for charcoal to be as much 
as M$ 20 700 per ha (US$ 8 333/ha) ; whereas stands allocated to fuelwood or 
pulpwood would yield a gross revenue of M$ 7 600 and 2 200 per ha 
respectively. As a comparison, inland forests were estimated to have a market 
value of M$ 11 500/ha after a rotation of 60 years. 

Ng (1987) compared the economic productivity of the Matang mangroves, 
which have been under continous forest management since the beginning of this 
century for charcoal and pole production, with the productivity of fishery and 
agriculture crop systems in comparable sites. According to his study, the 
management of the mangroves yielded a range of forestry and fishery products, 
of which the aggregate value was of the same order of magnitude as that 
realisable from a well -managed agricultural system. Also, after over 80 years 
of managed exploitation, there was no evidence of decline of overall 

Most reports mention only benefits derived from forestry and fishery 
products of the mangroves. Furthermore, since mangrove forests have so far 
only been managed sustainably in a few countries over a sufficiently long 
period to generate relevant data on costs and revenues, such data are 
extremely rare and in many cases have only localised value. 

There is thus a need to develop standard criteria for valuing intangible 
benefits from mangroves and to include these in order to arrive at discounted 
values per unit area, which can then be used as a basis for cost comparisons 
with other forms of land use. Utiless and until this aspect is clarified, 
through case studies such as the ones illustrated in Boxes 3.2 and 3.7, the 
value of the mangrove resource will not be realistically appraised to reflect 
both its tangible and intangible contributions to local , regional and national 


Using actual data 

costs fee Net Present Value (NFV) of forestry and fishery were estimated usittg (be faM^et^f p|Hl>idiiiim a 

social discount rate and a SO year planning horizon. 

id w amount of wood hai^estedx market vatoe - htrveiagoo*s. 

Commercial net benefits 

Subsistence net benefits were calculated using the actual amount of wod harvested x ttic slu^ow i^atee 
in the form of the price for inland or nsangtove firewood sold by ficeoced wood concessioners. 

Taking the speeds composiition of the mangrove area into account the weigjaed average Net IPrewat Value 
was estimated for each of the three main mangrove areas yielding the following result: 

Total Net 
Pisherv Met Benefits 

Value: $ M4417/IW 

Only in one of fee three areas was fee fisheries potential judged to be lully utilized and fee data are based 
on this area. 

Annual catch (commercial and subsistence): 3 026 tons 

Area of mangroves: 9 136 ha 

feus averaging 331fcg/ha equalling $ 864/ha/ in maifcet value amially. 

Taking harvesting costs ioto account gave the following rank: 
Net Present Vahie: $ 5 4/ha, or SJW/ta/yr 

This is assuming a proportionate decline in the fisheries. Wife only a 50% decline (as some of fee fish 
are not entirely dependent on the mangroves) the figure for fee NPV is $ 2 734/ha. 

Other Services: 

The value of mangroves for nutrient filtering hat been estimated using the fttertatto cost autfcod by 
Green (1983), who compared the costs of a conventional waste water treatment plant wife the use of 
oxidation ponds covering 32 ha of mangroves. An avenge annual benefit of $ 5 20/ka was obtained. 
This figure is however only valid for small areas of mangroves and as It represent* toe average aad not 
the marginal value it should be treated wife caution. 

The optfw vs^ and fee talrta^ 

and aii attempt to include these values was made by using tto eoMpMMtioo road 

fishing rights in Fiji caused by the reclanu^on of mangro^ has bn congtt^ 

Therecompeisumifi dttermiaedby anin^^ Urge 

vwiations in recompense sums were howver recorded <$ 49 - 4 45M^) according to fee ead w and the 

bargaining power of the owner of die filing rights. Using 1986 prices the following rewtts were 



$ 6(Wh* f or tofautrial we 

(My the iut lerek fii conpftnWe to die beaeto i 
2764 retpeotiveiy). 

It if ttuu 

Box 3.7: Sodo-economk value of nanfTOTCi in Fiji 



The following chapters review the information needs with regard to 
mangrove management, and describes in detail some of the tools used in 
obtaining information on the distribution and extent of mangrove forest 
resources, viz. remote sensing techniques, surveying, mapping, area 
computation and forest inventories. Emphasis is put on areas where these 
methods differ from conventional methods due to the special characteristics 
of mangroves and some prior knowledge of the subject is thus assumed. 


In order to develop an appropriate national or regional plan for the 
management and conservation of mangroves, a comprehensive data-base should be 
available, including information on the distribution and extent of mangrove 
areas, forest composition, actual and potential production and ecological 
factors which govern mangrove dynamics. Studying mangrove ecosystems for 
management purposes is not an easy task. Its complex nature and the different 
land-uses it may be allocated to, call for the concourse of expertise in 
various disciplines including Forestry, Ecology, Geomorphology, Aquaculture 
and Agriculture. 

As a first step towards integrated management of mangroves a survey of 
the mangrove areas should be undertaken. The kind of survey to be applied 
depends on various factors such as the size of the area to be covered, the 
type of vegetation, the purpose of the survey, the funds available, etc. In 
mangrove areas, the pressure on the resources (wood and non-wood products) and 
the land (agriculture, aquaculture, etc.) is increasing every day. 
Alternative uses of the forest land and the resources should be evaluated 
efficiently and quickly. The information needed for this evaluation includes 
the entire range of biological, physical and socio-economical data. One might 
be inclined to use a mult i- resource survey approach to gather data which will 
be used in the evaluation of the production trade-offs and conditions of 
resources. However, the application of such an approach in mangrove areas 
involves problems of different types such as: 

The identification of information needs is a complicated matter as each 
application field has particular information requirements, and 
different data collection techniques. This discordance makes the 
designing of a single survey to provide all the information required 
rather difficult. 

The requirements for management information of all resources are not 
the same in complexity, in detail and coverage. In mangrove areas, 
potential sites for aquaculture or agriculture may for instance not be 
present everywhere in the forest. 

Consequently, with different data requirements, multi -resource surveys 
could increase rather than decrease survey costs, and it might thus be 
advantageous to conduct a stratified or multiple-phase survey instead, where, 
based on the national/regional survey results, certain areas are allocated for 
ifere detailed and specific surveys. 


In the context of these guidelines, only forest related survey aspects 
will be dealt with. It should however be kept in mind that information from 
other disciplines in many instances is essential to the integrated management 
and utilization of the mangrove resources. 

For information on surveys covering other aspects of mangrove ecosystem 
management, such as geomorphology, fisheries resources, wildlife and ecology! 
please refer to more specialized textbooks and the literature cited in the 
relevant sections of Part I. For socio-economic surveys refer to Chapter 3. 
An excellent account of the legal and institutional issues with regard to 
integrated coastal area management, using a case study from Tanzania, is given 
by Young (1993) . 


The information needed for management planning purposes vary in 
importance, detail and complexity depending on the planning level. Whereas 
land use planning is often undertaken on a regional or national level and thus 
requires large scale surveys of areas of a considerable size, forest planning 
requires more detailed information on the forest types, the amount of wood 
available etc. 

An outline of some of the information needs at the different planning 
levels is illustrated in Table 4.1 on the following page. 

4.1.1 Land- use planning 

In order to formulate plans and adopt procedures to implement 
them, it is first necessary to evaluate and classify the land according 
to its various present uses and future potential. The principles and 
basic concepts of land evaluation are extensively discussed in the FAO 
forestry paper "Land evaluation for forestry" (FAO, 1984) . 

A land-use classification in mangroves should reflect the actual 
ground uses both for forestry and non- forest, land-based uses. In the 
forestry portion of the land, the information generally required for 
integrated management planning includes: 

The geographic distribution of mangroves and their extent: the 
area and the location of mangroves is probably the primary 
element to be acquired in a classification process. In fact, 
before a detailed classification of land uses within a mangrove 
area can start, the boundaries of the mangrove itself are to be 
identified. One must therefore be able to distinguish between 
mangrove formation and other vegetation. At this point, a clear 
definition of vegetation categories must be given in order to 
assure a consistent classification and to subsequently obtain 
reliable area figures. 

Mangrove forest resources and their potential supporting sites: 

The extent and distribution of productive and protective forest 
stands and areas where mangrove species could be introduced 
compose the most relevant information needed. 


Table 4.1: 

of iofonnrton needs * Affera* pbnrinf levels 






Information needed 

Relevant disciplines 

Land -use 

Long term 

Location and extent of 

Remote sensing (often 

Legal and 


forest and non -forest 

satellite images) 



Classification and 

issues 1 


Land tenure system 

Geology and 1 



Land-use planning 


Land- use potential a 



Estimation of forest 

mapping and forest 




Aquaculture 6 





Medium term 

Productive and non- 

Remote sensing 


management * 

(10 years) 

productive areas 

(aerial photography) 



Classification and 


Forest types 


Demographic and 






surveys 1 

Vplume fc growth estimates 

Forest inventories 


Regeneration statue 


Management objectives 

Mangrove management 

Manpower and equipment 





Demand forecasts 


Short term 

Wood volume 

Forest inventories 



(1 year) 

and maintenance 1 

Nursery and planting 

of roads, 


canal a and 




Thinning and other 

silvicultural treatments 

Harvesting systems 









Manpower and equipment 


Conservation and protection 


__ _^ B1lil __ B __ 





Regarding non-forest land-uses within mangroves, interests might 
be placed in the areas to be used for: 

Aquaculture: Areas to be alienated for fish ponds and shrimp 
farms ; 

Agriculture: Areas to be cultivated with rice and other 
agricultural crops; 

Mudflats, embankments and sand dunes, as well as areas reserved 
for construction, infrastructure and urban settlements; 

Waterways and drainage networks; 

Alluvial deposits. 

Protected areas and wildlife habitats 


The experience acquired in various countries where the management 
of mangroves has started, shows that most of the information above can 
best be presented in the form of small scale thematic maps (about 
1:250 000) . Maps at such a scale cover large areas, and at a national 
or regional planning level most of the items cited above can be 
represented with sufficient precision. Since the purpose of land-use 
planning is to allocate different lands to different utilization 
purposes, compatible with their suitability, the information required 
is mainly the land location and area by classes using sound and clear 
classification criteria. 

4.1.2 Forest planning 

Management planning 

The particular ecological conditions of mangrove formations, and 
the important socioeconomic value of the products and the services they 
provide, predestine these formations to some kind of multiple-use 
management. It is only by such an approach that conservation, 
production and recreational functions of mangroves can be fulfilled. 

Not all mangrove areas are equally productive or have the same 
potential. Allocation decisions should therefore take into account 
potentially compatible uses in order to allow a diversity of 
activities, prevent irreversible situations that may be caused by 
single use options, and ensure the integrity of the ecosystem. The 
objectives of managing mangroves for wood production as a primary use 
are briefly outlined below: 

Maintain continuous supplies of domestic and industrial wood 
products, such as poles, fuel -wood, posts, etc. 

Assure the regeneration of commercially valuable species and 
stands using adequate techniques. 

Conserve and intensify the protective function of the forest in 
designated areas and along river banks, estuaries, and all other 
marginal forest lands. 

The primary information required to achieve such objectives 
include a thorough knowledge of forest type composition and location, 
the volume of the growing stock, the growth rate and the regeneration 

Operational planning 

Operational planning deals with the techniques and strategies to 
be implemented in forest operations, such as for instance logging and 
drainage. Primary information needed includes the timber volume per 
species, size and quality. In addition, the knowledge of the precise 
timber location and the accessibility of the areas to be harvested, 
thinned or planted is essential. 

For forest planning purposes a more detailed forest inventory is 
thus needed as opposed to land-use planning, where a forest survey and 
a rough estimate of total volume in most cases would be adequate. 

4.1*3 Monitoring & evaluation 

The general concern over the progressive deterioration of 
tropical forests and the need for reliable information necessary for 
management decisions and conservation measures have led several 
countries to initiate national monitoring & evaluation programmes. 
Although the objectives of such programmes might be specific to various 
natural conditions inherent to differences from one country to another, 
they primarily aim at the assessment of forest cover changes over time. 

Guidelines and procedures for planning a monitoring & evaluation 
programme are dealt with in several reports and documents (FAD, 1985) . 
The purpose of including this section in the present guidelines is 
solely to indicate that such programmes refer to comparisons of forest 
conditions at two or more occasions, based on periodic surveys. They 
involve many activities common to other applications of forest surveys 
and remote sensing treated in later sections. These activities concern 
forest classification, mapping procedures, designing, planning and 
implementation of surveys. 


Effective management of forest resources calls for large amounts of 
current information. For most forest areas, primary information needs include 
the extent and distribution of forest cover and an assessment of woody 
biomass, vegetative production and forest conditions. Due to the particular 
forest structure, composition and difficult accessibility of mangrove forests, 
the task of collecting this information is often time consuming and therefore 
very expensive. 

The implementation of a survey, which can result in classification and 
mapping of the resources and the potentials of the area concerned should be 
the first step in obtaining the above information. In this respect, various 
remote sensing techniques have proved to be extremely valuable tools for rapid 
and relatively inexpensive collection of primary data, and a short description 
of their usefulness in mangrove areas is found in the following chapter. 

Chapter 6 deals with the aspects of planning and implementing mangrove 
surveys on different planning levels including classification of mangroves and 
presentation of the survey results in the form of maps. These surveys range 
from cartographic and multiple -phase surveys for land-use planning to more 
detailed surveys for forest planning purposes. 

The next chapter deals with the aspects of resource assessment and 
forest inventories conducted to obtain more detailed information on 
particularly the wood resources available in mangroves. 

Then follows Part ZV of these guidelines focusing on the more detailed 
information needs for mangrove management on a regional /district level, viz. 
silviculture i management practices, harvesting techniques, conservation issues 
and multiple-use of the resources on a sustainable level. 


The need for accurate information by forest managers and policy makers 
is crucial in mangroves, where absence of data in many countries has been a 
real obstacle to any management. This situation might be due to lack of 
knowledge of existing techniques, lack of necessary funds and skilled 
personnel, or merely because information obtained with remote sensing 
techniques has been reserved for non civilian use. 

When remote sensing techniques are to be used, any data analysis 
requires from the user a minimum knowledge of the subject for which these 
techniques are applied. He or she must also be aware of the limitations of 
these techniques with respect to practicality and accuracy as well as the 
quality and quantity of the information which can be extracted from air or 
space remote sensing data supports. 

The ability of remote sensors to distinguish between terrestrial 
features depends on various factors among which are the spectral 
characteristics of the objects on the ground and their morphology as well as 
the discrimination capabilities of the sensor used. 

Among the wide range of systems available and procedures developed in 
using remote sensing for identifying vegetation cover, its classification and 
mapping, one should use the technique or the combination of techniques which: 

Permit a rapid acquisition of data; 

Provide the required information ; 

Are cost effective. 

The major remote sensing systems which are recognized to have a 
veritable place in vegetation classification and mapping - particularly with 
regard to mangroves - include aerial photography, multi-spectral scanners 
(MSS) and radar. The principal advantages of each are: 

Aerial photography is simple to use, and lends itself to wide scale 

MSS provides .a wider range of information than other systems, and is 
suitable for automatic data processing and satellite applications. 

Radar presents all weather capability, which is particularly useful in 
cloudy areas such as the tropical zones, and is also suitable for 
regular monitoring; 

A summary description of the technical characteristics of various remote 
sensing sources is presented in Appendix 1 in the back of this paper. 


In mangrove forests the application of remote sensing can be envisaged 
in three planning levels: national level, regional/local management level and 
operational level, each one aiming at specific goals. 


To these planning levels one might associate three levels of land 
classification which can be identified as : 

Mangrove land-uses; 

Foreet eite claeeification; 

Forest type mapping. 

The level of detail may also determine the criteria to be considered for 
classification. Geographical and ecological criteria for example are often 
adopted at the most general level of investigation, whereas functional 
criteria are more appropriate at the most detailed level. In case of 
extensive surveys using remote sensing techniques, physiographical criteria 
are important. 

The principal elements on which the choice of a sensor is based - for 
each one of these situations - are data characteristics which in essence 
determine the quality and amount of information, and the site of the area 
concerned by the survey. Most of the currently applied remote sensing systems 
- using these criteria - can tentatively be ordered in a hierarchical manner, 
to fit more or less the planning levels mentioned above. An exact fit is 
obviously not attainable due to the overlapping ground resolution ranges 
between sensors but it gives a somewhat useful indication on the ability of 
remote sensors to meet the requirements of the type, quality and quantity of 
the information really needed in a given situation. Diagram 5.1 below 
indicates approximate resolution requirements associated with data survey 
levels and major sensors used as might be applied to mangrove forest surveys. 



0~i 10 100 

Physiographic features 
Pattern of human activities 
Drainage pattern 


All of the above items 

Land use classification 

Broad vegetation classification 

All of the above items 
Forest type identification 
Forest measurements 

Aerial photography 

Spot imagery 

Landsat Thematic Mapper 

Landsat Multi Spectral Scanner 

Side Looking Airborne Radar 







5.1,1 Application of aerial photography to mangrove areas 

Panchromatic, colour and infrared colour aerial photography have 
been the basis for various mangrove forest surveys and inventories. At 
small scales, they can be used in reconnaissance surveys , or for broad 
forest type classification in extensive areas. Prom medium scale photo 
coverage, one may obtain detailed forest stratification - based on tree 
cover sizes or forest stand heights- and land use identification. 
Large scale photography is a valuable support for forest stand 
measurements. In many cases it can be used as "ground truth" in 
surveys where small and large imagery are combined. Low level aerial 
reconnaissance survey using a light aircraft and small format cameras 
may play a crucial role in this respect. Through such flights, it is 
possible to gather accurate information about mangrove forest 
conditions, denuded areas, fish ponds, plantations and other land uses 
and coastal features. 

From examples of aerial photography applications to mangroves, 
with the objective of spscits recognition, Rollet (1974) showed that on 
1:33 000 scale panchromatic images , pure stands of Avicennia and 
Rhizophora can be distinguished. Species separation is, however, more 
difficult in mixed species stands. On 1:20 000 scale panchromatic 
films, Avicennia could very well be separated from Rhizophora stands 
due to their grey tone and coarser texture (see Figures 5.1 and 5.2). 
Hamilton and Snedaker (1964) indicated that on 1:25 000 scale, black 
and white aerial photography, the mangrove genera were easily 
recognized by crown size, tone and relative height. 

The introduction of infrared photography in combination with 
panchromatic films improves the ability to recognize mangrove species 
a great deal (Rollet, 1974). This is illustrated in Figure 5.3. The 
author also pointed out that crown texture is a major element to take 
into account for improving species recognition accuracy based on colour 

With regard to Bangrovs land utilisation, black and white aerial 
photography proved to be useful in identifying and mapping newly 
accreted lands, based on tonal differences of surface dryness and 
moisture conditions. It has also been used to classify new mangrove 
plantations according to tree height and stand density (Rahman et al., 
1986) . In another study for the assessment of location and extent of 
shrimp and fish ponds, Shahid and Pramanik (1986) pointed out that 
black and white (1:30 000) and infrared coloured aerial photography 
(1:50 000) revealed to be valuable. 

In mangrove areas where accessibility is a real problem, low 
altitude aerial photography is a useful tool which can be used 
successfully in conjunction with limited ground data to document forest 
canopy and other non-forest land-uses. 

This type of photography, usually at a large scale, contains 
precious information which can be used as "ground truth" either to 
correct misinterpretation or as reference data for construction of a 
photo- interpretation key. Being very versatile, it can also be 
combined with satellite imagery, to document environmental conditions 
which prevail at the time when the photographs were being taken. 

Figure SJ 

Mb pur : Pure stand of large Laguncularia 
(Mangle bianco) 

Figure 5.2 

11 : Pure stand of large Laguncularia 

trees with regeneration ofRhizophora 
and some Avicennia 

IS bis : Large Laguncularia trees 
(Mangle bianco) 

15 ter : Low Avicennia stand 

11 bis : Patches of semi-deciduous 

11 ter : Dominant Laguncularia 

(mangle bianco) behind 

12 : Pure low Avicennia with many large 

rotten Laguncularia trunks 

13 : Pure stand of Laguncularia 

(25m high and 30 cm DBH) 

5.1tnd5J: Aerial photo showing manfrove vefetation types in Mwico. 
Scale: 1:20000. (Source: Rollet, 1974) 


Fig. 5.3: IR colour aerial photos showing mangrove vegetation types (Mexico). 

(Source: Rollet, 1974) 

Co : Conocarpus 

Mb : Large Laguncularia (Mangle bianco) 

Me : Laguncularia thicket (Mangle chino) 

MSD: Semi-deciduous thicket 
Av : Avicennia sp. 
Rh : Rhizophora sp. 


5.1.2 Application of satellite imagery to mangrove areas 

In tropical countries where vegetation survey programmes have 
started, classification and land-use mapping include mangroves as one 
of the vegetation classes within the whole forest land group. Few 
studies, however, have been conducted specifically to gather 
information on mangroves. Landsat imagery using the Multi Spectral 
Scanner (MSS) has been the main source of information and both visual 
and digital data analysis have been applied with varying degree of 
success. On photographic images, mangrove vegetation is often 
characterized by a smooth texture due to small tree crowns and 
relatively dense stands. The association of mangrove vegetation with 
estuaries is also a valuable element for its identification. 

Numerous studies have reported that mangrove forests can be 
easily separated from other vegetation formations on satellite images 
(Charuppat, 1983; Chaudhury, 1983, 1985, 1986; Silapathong, 1983). 
When image interpretation is carried out on black and white and diazo 
prints for instance, Ishaq-Mirza et al. (1986) indicated that mangrove 
vegetation could be classified into dense, normal and sparse 
vegetation, based only on tonal differences. Also, at the original 
scale of 1:1 000 000, diazo colour prints may give better tonal 
contrast than black and white images at a scale of 1:250 000. 

Mangrove cover types classification based on species and density 
has also been attempted using Landsat MSS data. Not all species could 
be separated and the persisting confusion between cover types indicates 
that cover type maps from Landsat data would not be successful in mixed 
mangrove forests. However, natural mangroves are easily distinguished 
from planted areas (Chaudhury, 1986) . 

From the few Return Beam Vidicon images available of mangrove 
areas, the same author pointed out that most rivers, canals and creeks 
could be identified and enlarged images (1:250 000) have been used to 
compile a map of the Sunderbans in Bangladesh. In addition to the 
precise location of mangrove features in this extensive formation, this 
map is also used for navigation purposes. The advantages of this system 
is that it can produce more planimetrically correct results than MSS 
images . 

More detailed classifications of MSS data within mangrove 
formations have been tested and most results showed that visual 
interpretation of low resolution satellite imagery should be limited to 
broad forest and land-use classification. 

Digital classification of satellite data has also been considered 
in numerous vegetation studies including mangroves. Various attempts 
have been made for mangrove surveys, land-use identification and even 
forest type mapping. 

In terms of classification accuracy, computer processing of 
satellite imagery was found on various occasions to be superior to 
visual classification. By using image data in computer compatible 
format, greater flexibility in image processing can be achieved. 
Moreover, no radiometric details are lost as it happens through 
photographic processing. 


Fig. 5.4: A partial view of a False Colour Composite obtained from combinations of ratios 5/4, 5/7 and 

6/5 of Landsat MSS bands. The mangrove appears as orange-red, marsh vegetation as 
reddish black, shrubs as yellow, crops as bright red and saline areas as white. 
(Gulf of Kachc Landsat Data, India, 1982) 
(Courtesy: Nayak) 

Fig. 5.5: An FCC of the same area as above, obtained from combinations of band 5 and ratios of 5- 

7/5+7. Mangroves appear as yellow, marsh as greenish black, swamp vegetation as red and 
crop-barren land as whitish blue. 
(Courtesy: Nayak) 

Figures 5.4 and 5.5: Examples of Landsat MSS imagery of mangrove areas. 


From the results obtained, it also appears that spectral data 
transformations are essential as they bring out differences among land 
cover categories* The most successful techniques include contrast 
stretching, principal component analysis, and band ratioing. (Refer to 
Appendix A2 for more details on these techniques) . Ratios of Landsat 
band 5 and 7 and 5-7/5+7 performed best. Nayak et al. (1985, 1986) 
mentioned that ratios based on these combinations greatly enhanced 
subtle differences between mangroves, swamp and marsh vegetation (see 
Figures 5*4 and 5.5) . Mangrove vegetation could also be separated from 
marsh vegetation on false colour composites issued from principal 
component transformed data. Digitally processed Landsat data was also 
found to give useful information on the location of mudflats, sandy 
areas and newly accreted lands. 

The application of image smoothing to mangrove classification 
resulted in the distinction of two different classes: dense and open 
vegetation. An .even better definition of forest boundaries was 
obtained when the central pixel value of the blocks was replaced by the 
median of block values instead of their mean. 

In addition, in order to obtain more accurate outputs, sufficient 
"training areas" should be established and scattered over all cover 
types of the mangrove area, where ground truth data can be established. 
Since mangrove types in most cases follow a more or less distinctive 
zonation pattern, such training areas can often be uniformly 
distributed along a transect line extending from the sea shore to the 
main land. 

As to major mangrove sp*ci*s identification, and forest type 
(species-density) separation, supervised and unsupervised 
classifications of Landsat data were applied without satisfactory 
outputs. Because of the inherent limitations of Landsat MSS - due to 
its low spatial resolution - resulting in unsuccessful identification 
of individual tree species, attempts have been made, based on SPOT 
simulated data. A few recent applications to mangrove classification 
showed that although a complete separation of individual species could 
not be achieved, substantial improvements could be expected in both 
forest and other non-forest mangrove classification (Abdus Shahid, 
1985; Berenger, 1985; Hossein, 1985) . From SPOT images, enlargements 
are also used and scales of 1:100 000 and even 1:50 000 are feasible. 

As it appears in Figure 5.6, Lantieri (1986) indicates that 
mangrove forest could be mapped accurately (90% accuracy at 1:50 000 
scale) , but major species separation is possible only in case of large, 
pure stands. The author also pointed out that when contrast 
enhancement or band decorrelation is performed, colour composite 
interpretation shows that some mangrove areas could be separated by 
their density. Other encouraging results were achieved by Blasco et 
al. (1986) in classification of a coastal tropical zone including major 
species of mangroves (See Figures 5.7 and 5.8). 


Fif.5.C: False colour composite of a SPOT li 

Scale: 1:50000 Resolution: 10m 
(Coastal zone at Ngomeni, Kenya) 
(Source: Lantieri, 1986) 


Fig. 5.7: Colour composite of SPOT image and coirespcmding sketch m 

Resolution: 20 m (Bangladesh) 
(Source: Blascotf a/., 1986) 

Legend : 

1: Natural den0e mangrove (mainly Sonneratia) 

2: Natural open mangrove (deciduous Sonneratia) 

3&4: Plantations 

5: Nursery 

A: Brackish water 

C: Sandbar 

11: Grass 

12: Jfypa f rut leans 

13: Sonneratia apelata 

14: Exooecaria agrallocha 

15: Predominantly Her i tier A fames 

(Soiree; Btocorf a/., 


From SPOT simulated data, Loubersac (1983) and Hosscin (1985) 
reported that it was possible to achieve accurate classification of 
mangroves, by using a prior stratification before analyzing multi- 
spectral images of each unit. With this procedure the variance is 
reduced, and the results obtained can optimized by statistical 

Based on experimental results obtained from various applications 
of satellite imagery to mangrove mapping, it appears that no procedure 
of interpretation is absolute. It is up to the image analyst to 
examine all combinations of image products at hand in order to find the 
one which leads to the best output as, for the extraction of each 
category of interest, different enhancement procedures might be 
necessary. With all the systems available, the success in converting 
image data into useful information still depends primarily on the depth 
of the user's knowledge of the specialized subject for which these 
techniques are applied, and the ability to comprehend the effects 
involved in interpretation analysis. 

5.1.3 Application of radar imagery to mangrove areas 

In many tropical countries, persistent cloud cover during long 
periods prevents the acquisition of cloud free images with photographic 
or MSS sensors. In spite of their proven performances, their 
limitations are their incapability of cloud cover penetration and their 
sensitivity to atmospheric disturbances (See Figures 5.9 and 5.10). 
One approach to overcome the problem is to use synthetic aperture radar 

Among the various applications of SAR to forest resources 
studies, mapping of coastal forests and mangroves is a case where the 
system has been the most successful (Imhoff and Vermilion, 1986) , This 
sensor can be used for surveying extensive areas -and can be useful in 
the preparation of small scale (1:200 000-1:500 000) regional or 
reconnaissance maps of vegetation types. These maps can subsequently 
be used to select priority areas for more detailed remote sensing and 
ground survey studies. 

On radar imagery, forts t species cannot bs identified. All 

information required for forest management, such as species 
composition, forest structure and stand description, should be obtained 
from ground and/or air reconnaissance surveys. In addition, all water 
bodies display the same dark tone, and sand bars .and mud banks near 
estuaries must be above water level to be recorded on the image. In 
spite of this limitation, radar imagery provides a good impression of 
the physiographic conditions of the terrain, and as a result, in low 
land areas such as mangroves, the drainage pattern is clearly visible. 
Furthermore, inundated zones can easily be separated from dry land 
forest (Sicco-Smit, 1975) . This is illustrated in Figure 5.11, which is 
a partial view of an image showing a coastal tropical area. 


Hg. 5.9: Air photo showing doud cover o?er a mangrove area 

(Courtesy: Sicco-Smit) 


Pig. 5.10: AradarlmafeoftbecoMUJzooeofColoinbta 

(The area covered by (he air photo in fig. 5.9 is indicated for comparison) 
(Courtesy: Sicco-Smit) 


fig. 5*11: A radar image showing mangrove formations 

Scale: 1:220000 (Colombia) 
(Courtesy: Sicco-Smit) 


Bank above water level 

Mangrove forest 

Swamp, salt or brackish water 

Limit between mangrove vegetation and 

fresh water swamp forest 

Swamp with herbaceous vegetation 

Low swamp forest 

Salt water of estuary 

Tidal creek 

Old meander 

5.1.4 Comparison between major sensors 

In order to choose the sensor and technique which provide the 
information needed, the fundamental question that should be asked is 
what kind of data and with what precision it can be obtained through a 
given technique, in order to meet the planner or manager's information 
needs? To answer this question, the user should be aware of the 
technical performances of different sensors, i.e. their ability to 
discriminate between various terrestrial features according to the 
resolution and the requirements which are appropriate to the type of 
information he/she seeks. In addition to the kind of information 
needed for planning purposes, the choice of remote sensing systems 
depends also, and to a great extend, on the size of the area to be 
surveyed and on the funds available. 

In a comparative study of interpreted aerial panchromatic 
photography and Shuttle Imaging Radar (SIR-A) images, Castellu found 
the results presented in Table 5.1 below. 

Table 5.1 : Comparative photo-interpretation of panchromatic photography and SIR-A images 
(Partial results from Castellu, 1985) 

Type of 


SIR-A images 





Small dense crowns 
dark tones, 60% of 
the mangrove area 

As Ml but with 
lighter tones 

Scattered trees on 
coast, light tones; 
pioneer trees. 

Very fine granularity 
salt and pepper (*) 

Confused with Ml 

Not identified 




No granularity 
Medium to dark 
grey tones; 

As Nl but less 

Dark grey tone 
Slight granularity 

As Nl but lighter (*) 




Very fine granulation, 
light tones 
low vegetation 

No granularity 
uniform slightly 
mottled tones 
low vegetation 

White narrow coastal 
strips (*) 

No identification 
(Newly created areas) 

(*) : Good identification 
(**): Very good identification. 


In order to improve the interpret ability of the image , as it is 
illustrated in Figures 5.12 and 5.13, Imhoff and Vermilion (1986) 
indicate that contrast enhancement techniques, filtering, and density 
slicing of the radar image can also be used. 

Below a summary of relative advantages and disadvantages of 
aerial photography, MSS and radar is listed, as reported by Roberts 
(1975) . 

Table 5*2: Relative advantages of nutfor sensors 

Advantages of 
MSS vs. Aerial Photography 

of MSS vs Aerial Photography 

Wider range of spectral 


Easier calibration of signals; 

Good registration between 

spectral bands; 

More suitable for digital 

conversion and automatic 


More suitable for satellite 

application and more suitable 

for regular multitemporal 


Poorer resolution; 
Less suitable for local 

Advantages of 
MSS vs Radar Imagery 

Disadvantages of 
MSS vs Radar Imagery 

Greater variety of information 

allowing a greater ability to 

distinguish between different 

vegetation conditions; 

Fewer problems of geometric 


Better resolution possible. 

Much greater interference from 
cloud and haze consequently much 
less all weather capability; 
Inability to obtain imagery at 
night except from thermal infra 
red channels; 

Reliance on variable levels and 
quality of sources radiations 
instead of constant output for 

Less ability to obtain 
information from beneath 
surface of vegetation. 


Fig. 5.12: A black and white Shuttle Imaging Radar (SIR-B) image 

(Notice the salt and pepper appearance of the image before the filtering) 


(Courtesy: SPARRSO - Irahoff) 

Fl|. 5.13: Enhanced colour composite of a SIR-B radar Image 

(The same area as in Fig. S.12 above) 
(Courtesy: SPARRSO - Imhoff) 



The objectives of a forest survey are to obtain reliable and up-to-date 
information about the quantity and quality of the forest resources. The type 
and quantity of information concerned by a forest survey and the means to use 
for its implementation depend on the area to be surveyed and the funds 
available, and differ according to whether the information is required for 
land-use planning on a national or regional level, for forest land management 
on a district or forest level or for operational planning* 

The degree of detail of a forest classification based on a survey 
depends on the information level required and the remote sensing data source 
used* Information derived from large to medium scale imagery can be used at 
all levels of planning* However, due to cost constraints at these scales, 
such imagery is commonly only used as a basis for operational planning and 
forest measurements where more detailed information is required, while land- 
use data can be obtained from medium to high altitude photography and orbiting 
satellite imagery at a small scale* 

The use of satellite remote sensing in mangrove ecosystem studies is 
relatively recent, but significant advantages of satellite imagery have been 
identified in several applications. They include: 

The feasibility of large scale surveys at lower cost per unit area. 
This is a fundamental element to be considered as the land with 
mangroves and the coastal zones can be extremely large* 

The possibility of using computer processing to improve the information 
value * 

In mangrove ecosystems which are rapidly changing, the satellite 
ability of repetitive scanning makes monitoring changes much easier. 

In extensive areas, land-use classification, which does not require 
detailed information, can adequately be carried out through space and/or 
aerial surveys, and the location of the lands may be properly presented on 
maps or aerial photography of a small scale. Area figures and a general 
description can also be provided to complete the information* 

In spite of the advantages that satellites present, in terms of rapidity 
and regularity of providing data, the actual low spatial resolution of their 
imagery does not meet all the information requirements needed for management 
planning at district or forest levels. Aerial photography at medium and large 
scale is necessary for terrain data acquisition such as topography, 
physiognomy and accessibility* Information on stand and tree characteristics 
can also, to some extent, be obtained from large scale aerial photography but 
reconnaissance field trips are always necessary to complete and correct the 
data extracted from remote sensing imagery. 

Considering the cost, the time constraints and logistics associated wi~h 
the difficult task of surveying mangrove forests in the field, either to 
gather new information or to support a monitoring program, one should take 
advantage of both space and airborne remote sensing systems. 


A combination of satellite and aerial (conventional and wall format) 
photography can successfully be utilized in mangrove areas. It must however, 
be pointed out, that field surveys should be considered as an integral part 
of any land or forest survey at any level. In this regard it should be 
recalled that, the mangrove environment is very different from the inland 
forest from a logistical point of view, and special care should be taken for 
planning field operations (transportation means, supplies etc.). 


Management decisions, at whatever level they have to be taken, require 
the development of a classification system subdividing terrain and vegetation 
features into homogeneous categories, made up of similar items, based on 
environmental and physical characteristics. Such a classification system 
should be an integral part of any forest survey. 

Remote sensing plays an essential role in the establishment of 
classification systems, because it permits forest types and sites to be 
identified and mapped. A classification system is also the basis for forest 
stratification and is essential in the preparation of forest maps required in 
the planning and implementation of more detailed surveys and forest 
inventories, as well as in the preparation and execution of forest management 
plans. In the difficult natural working conditions characterizing mangrove 
forests, such tasks are hardly feasible without detailed maps. 

A general scheme for a mangrove classification system is presented in 
Diagram 6.1. As can be seen, the first step in a mangrove forest 
classification is the distinction between forested and non-forested lands. 
Forest recognition may not be as simple as it sounds, particularly in areas 
where mangrove stands have been subjected to advanced degradation. Much care 
should thus be taken, in order to establish- an appropriate definition of 
classes before the survey begins. 

Non-forest areas occurring within the forest boundaries should also be 
classified according to some definite system based on the physical nature of 
the lands but also, whenever possible, according to the use to which the land 
will be devoted. Non- forested land may be : 

Agricultural lands and saltponde; 

Lands presently used for or to be allocated to aquaculture; 

Urban and mining areas, infrastructure etc. 

Mangrove forest areas can be further divided into : 

Productive forests: comprised of well established and developed stands, 
where regeneration can be assured and, 

Non-productive forests: areas which presently do not have a productive 
forest cover. This includes degraded areas which might be turned ilxo 
productive stands (potentially productive areas), and lands suitable 
for protection of wildlife species and vegetation types where felling 
and extraction of wood should be avoided (protection areas). The 
latter also includes areas, which it is necessary to keep under 
permanent forest cover in order to counteract erosion and the resulting 
siltation of rivers and estuaries. 









co < 
w w 




















2 CO 
H W 

w O 
fe U 


H H 



o oi 




Forwt typ classification is usually based on aerial photography. For 

this purpose, medium scale (1:10 000 - 1:30 000) will in most cases be 
appropriate to identify vegetation types and assess their pictorial 
characteristics. In regions where few forest tree species are present, black 
and white aerial photographs can be satisfactory. When forest stands contain 
numerous species, colour and IR coloured imagery may be best. 

Colour photography provides a significant increase in accuracy in 
spftcias identification, and the increase in accuracy contributes in reducing 
survey costs. Species identification is not only accurate on colour film but 
interpretation results are achieved more quickly. The difficulty of species 
identification is increased when small scale photography is used, (as is the 
case in many tropical studies) , and also due to the lack of good standard 
descriptions of tree shapes on aerial photography. Field reconnaissance 
surveys should, in any case be undertaken to complement and rectify the photo 

Forest stands can also be identified using aerial photographs, but again 
field reconnaissance surveys as well as old maps, records and previous working 
plans are essential inputs to ensure an accurate classification. 

Photo- interpretation procedures for visual and computer-aided 
classification of mangroves using aerial photographs and satellite imagery 
(including radar) are described in Appendix A. 2. 


Forest survey designs are numerous. In planning a forest survey, one 
should select the most suitable for the conditions which prevail. The survey 
may be designed to cover a whole country or a province for a global resources 
assessment; to support a forest management plan; or to evaluate timber volume 
production in small size units. At each one of these types of survey, the 
information to be provided varies in detail and accuracy. 

A particular aspect to consider in mangroves is the change in area. 
This can be due to accreted land formed by coastal or riverine deposition, and 
may be used for new plantations, or a decrease in land area caused by chronic 
erosion due to a change in coastal currents. Table 6.1 gives an indication 
of the relative importance of a few survey elements. 

Table 6.1: Relative importance of forest survey elements in mangrove forests 


Area Acceaaibility Transportation Volume Growth 
itimata facilities alternation a a ae lament 

Land Soil stability 
tenure erosion/accretion 




Port at 






X : Little intereat 1 
XX : Nediwa priority 
XXX : Vary important 


At any level of information, the forest survey should be well prepared. 
In the preparation phase, decisions are to be taken on: 

The type of information needed; 

The survey method to apply in order to acquire the data needed with the 
least cost possible. 

At the same time, it is imperative to conduct a thorough study of all 
existing documents and data available about the area concerned by the survey. 
Such documents include past survey reports, maps, research papers, etc. 

6.2.1 National level surveys 

In a national level survey, the primary objective is to provide 
data which can serve for decision making on national (or regional) 
forest policy, and in the implementation of global development plans 
irrespective of whether the forest sector is already developed or not. 
In mangrove areas, a survey of this type is one where the main interest 
will be placed on the knowledge of the area extent of mangrove 
vegetation, its distribution and a broad classification of lands both 
for forest and non-forest uses. A few alternatives are discussed in 
the following sections. 

Cartographic survey 

A cartographic survey may be defined as a survey which is 
primarily based on image interpretation, and the objectives of which 
are often limited to the production of thematic maps. Over extensive 
areas, small scale aerial photography and/or satellite imagery are 
suitable data sources for such a task. Map compilation requires a 100% 
image coverage, and all images are interpreted following a 
classification scheme established prior to interpretation. 

The low resolution of satellite images does not allow detailed 
classification of all forest types. They can be used for 
classification of evergreen and mangrove forests only, but the 
identification of mangroves on MSS imagery is facilitated by their 
location along estuaries. On colour composites, their dark red colour 
is also very indicative. To increase the accuracy classification, 
combinations of spectral bands are used. Landsat band 5, for instance, 
is interpreted for mangrove forest delineation, various land-use types 
and road network identification. Band 7 is mainly used for identifying 
mudflats, coastal shorelines and water bodies such as rivers and 
reservoirs. Prints are interpreted separately but their comparison 
with each other and with colour composites is useful. 

Maps and mosaics, such as those described and illustrated in 
Station 6.3 and Appendix A3 are considered the main final outputs of 
the survey. 

The evaluation of map accuracy and correction of 

misinterpretation can be achieved by means of a limited ground check in 

the field. Low flying plane missions could also be successfully made 

: to raptdly check photo- interpretation and reduce the costly ground 



The procedure of ground checking consist* of drawing a number of 
ample units on images, locate them in the field and check their 
classification. Since the purpose of the sampling in this kind of 
survey is primarily to assess image classification and mapping 
accuracy, the easiest technique is to use a regular grid with the dots 
representing the location of the samples. Image and field 
classifications are then compared, using error matrix procedures 
discussed in Appendix A. 3. Stratified sampling may be applied by 
drawing an independent sample in each stratum, or following the 
selection technique presented in Appendix A. 3. 

Diagram 6.2 presents the sequence of activities in a cartographic 
survey based on a) aerial photography and b) satellite imagery 

The combination of aerial photography and satellite imagery can 
also be successfully be applied, as presented in Case Study 1. 
























* : In cast of visual sat* U it* inaga intarpratation, this stap ia skippad. 

Diagram 6.2: Sequence of a cartographk surrey baaed on (a) aerial photos and 

(b) Satellite images. 

Multi-phaae surveys 

In this type of survey, mapping may or may not be required. Data 
are collected on sample plots, which are either randomly or 
systematically laid out over the image. The approach of using 
multilevel data is designed for the application of aerial photography 
(black and white or colour) in combination with satellite imagery in 
the form of black and white prints and/or colour composites. 

The advantage of the method resides in combining the positive 
aspects of satellite images (the possibility of having a synoptic view 
of the' area of concern at low cost) , together with that of aerial 
photography which permits more details to be mapped. Aerial 
photography is used in this process, to complement, adjust and check 
the interpretation obtained from satellite images. In view of keeping 
the survey cost at low values, photo- interpret at ion of conventional 
photography is restricted to sample areas only. 

The application of the procedure in forest surveys results in a 
multi-phase sampling design. With such a design, both data bases are 
used to produce classification statistics with a higher accuracy 
compared to a classification based on satellite image alone and at a 
lower cost than if aerial photography were the main source of 



It is also here recommended to include in the process of 
interpretation any secondary information such as existing maps or any 
other documents, which may serve as a guide in the image analysis. 
Frequent reconnaissance field trips should likewise be undertaken while 
the interpretation is being carried out to enable the image analyst to 
become familiarized with the area of study and the objects to be 
identified and delineated. 

A double phase sampling design is a particular case of the 
multi-level survey scheme mentioned above. A brief account on this 
design is given in Appendix 4. The first phase of its application 
involves aerial photography. Satellite imagery can also be used but 
because of the low resolution, it might be very tedious or even 
impossible to correctly locate the sub-sample units on the ground. The 
second phase involves classification in the field. This technique can 
also be successfully applied when the area concerned by the survey is 
only partially covered with aerial photography, which may have been 
flown in independent strips, i.e. without overlap. Although mapping of 
the results is not always required, the procedure is easier if an 
existing map or a mosaic is used. 

The application of a three-phase stapling to mangrove forests 
using satellite imagery, aerial photography and field sampling as t.he 
3 phases is illustrated with an example in Appendix 4. 

Such mult i -phase sampling methods are often used to obtain an 
estimate of the area covered with mangroves and also for a 
classification into different forest types. The same technique can be 
successfully applied in resource assessments and forest inventories as 
described in Chapter 7 and Case Study 3. 

6.2.2 Forest management surveys 

In mangroves, as in other forests, the increasing consumption and 
production demands require from managers to answer the questions : 
What, Where, and How much, through analysis of data acquired either 
from ground survey, aerial survey or both. The information wanted is 
in most cases the locations and distribution of resources, tree 
species, sizes and quality, growth and site quality. Although all this 
information cannot be obtained through remote sensing, procedures 
combining statistical analysis with limited ground surveys could be 
successfully applied. 

While large scale (national) surveys concern extensive areas, 
forest management surveys are usually restricted to forests in a 
district or smaller unit. The information a forest manager expects 
from a management survey refers mainly to: 

An accurate area estimate; 

Classification of the forest into cover types and their 

Evaluation of the growing stock; 

Regeneration assessment; 

Assessment of tree and forest stand growth; 

Methods for estimation of mangrove areas are briefly described in 
Station 6.4 of this chapter. 

Whereas classification into forest types can be successfully 
achieved using aerial photography of an appropriate scale combined with 
limited ground checks, the evaluation of the growing stock and the 
regeneration often depend on more intensive field enumerations and 
ground surveys. This is also, and especially, the case when an 
assessment of the tree and forest stand growth is needed. In these 
cases the surveys are often combined with information gathered through 
forest inventories described in the following chapter and illustrated 
in Cast Study 2. 

Forest tvoe delineation 

Prior to field enumeration, a precise classification of the 
forest into cover types should be completed. The goal of forest 
classification is two-fold: first, it is essential for stratification, 
which can permit a reduction of field enumeration, and second, it is 
necessary in the preparation of management maps. This classification 
is, as described in Section 6.1, most easily carried out using aerial 
photographs . 

When planning the aerial photo coverage of mangrove areas, 
particular attention should be given to the season of the year and the 
time of the day to ensure an accurate interpretation. For vegetation 
mapping and interpretation, the best season is when differences between 
trees and vegetation classes can be easily detected. 


In mangrove areas, this may correspond to the end of the rainy 
season. Also, photography should be taken when the sky is free of 
clouds and dust. 

At the latitudes where mangroves are located, the best time for 
photography is between 9:00 am and 3:00 pm. However, at noon the high 
elevation of the sun may cause specular reflection in some spots 
especially when wide angle cameras are used. Moreover, consideration 
should also be given to tidal levels and amplitude as they may affect 
the photo- interpretation results. A few of the advantages (+) and 
drawbacks (-) associated with low and high tides are indicated below. 

Table 6.2: 

The I 

of tidal level on aerial photography coverage 

Low tides 

High tides 

+ Mudflats in lower littoral 

zones are visible 
+ Detection of areas with 

stagnant water made easier 

- Not adequate for studying 
inundation levels 

- Cannot detect interdis tri- 
butary canals, which can 
be useful in planning 
field survey (access) . 

+ Zones of maximum inundation 

are visible 
+ Most canals can be detected 

- May cause bias in tree 
height measurement on photos 

- Detection of rivers and canal 
systems in degraded or non- 
forested areas not possible. 

It must also be pointed out that for forest classification, 
highly qualified photo -interpreters who are familiar with the mangrove 
environment are required. 

Valuable additional aids to photo -interpretation, besides photo- 
interpretation keys, include terrestrial photography revealing the 
ecological sites of each forest type, stereogrammes and type 

With regard to forest typing in mangroves, a principal element of 
classification is often the potential commercial value of timber and 
other wood products. A separation between productive and non- 
productive forests will result in a more rational utilization of the 
means used in the survey. In areas of poor wood production potential 
for instance, a visual description of the vegetation might be 
sufficient whereas in productive stands a more detailed forest 
inventory including information on species composition and volume is 

According to Diagram 6.1, stands where the forest presents a 
potential value of timber and other wood products, can further be 
subdivided into more refined classes such as forest types. The latter 
can be defined as being groups of trees having the same 
characteristics, growing in the same conditions, and having the same 


Their separation can be achieved using easily distinguishable 
criteria such as: 


tree height 

stand density 

With regard to composition, genera can be used where the 
distinction between individual mangrove species is not feasible. In 
addition, ecological factors such as texture and condition of the soil, 
and inundation classes can be also incorporated to achieve a better 
classification of forest sites. 

The age is also an important classification element to consider, 
but in mangrove forests, unless the stands are under intensive 
management, such as for instance in the Matang Mangrove Reserve (Perak, 
Malaysia) , the age will usually not be known. 

Tree height is also a useful criterion in forest stand 
classification. Its usefulness resides in the fact that tree height can 
be measured with acceptable accuracy on suitable aerial photographs. 
In mangrove forests, stand height can be divided for instance into 
three or four classes, allowing a sufficient description of stand 
development such as 

Stand height 

- 9m 

10 - 19 m 

> 20 m 


Young stand 
Old stand 

Height classes may vary in amplitude from one species to another, 
but their number, should be kept reasonably low (3 to 4) . Due to 
variations in species and growing conditions between countries, and 
even within a country, standard height classes are not recommended 
here. The classes should be based on local observations and 
information needs. 

It should be recalled here that, when tree height is determined 
from aerial photography by means of parallax measurements, it may be 
necessary to make adjustments to correct for tidal fluctuation. 

Forest density is a measure of stocking in a forest. It is also 
a valuable classification factor, which is frequently used, because it 
is highly correlated with volume. Moreover, it can be directly 
expressed using crown closure measurements on aerial photographs. Like 
stand height, forest density can also be subdivided into several 
classes for instance as follows: 


Crown closure 
in percent 


10 - 40 
40 - 70 


> 70 


Again, it is not possible to generalize with regard to the 
number of classes and their limits. The classes should reflect the 
occurrence of the local species and their characteristics. 

Most of the criteria mentioned above (and which are used in 
forest type classification) can also be the basis for the establishment 
of photo- interpret at ion keys. For their construction, it is first 
necessary to determine the photographic characteristics of the 
different species of mangroves under various conditions. Then, from 
observations of disseminated ground plots covering various vegetation 
types, sites, growth stages and degree of utilization, one can 
establish physical links with actual ground conditions. 

The simplest way of providing the photo- interpreter with useful 
reference material is probably by using annotated stereogrammes which 
show various objects to be identified. An example of such a 
stereogramme is presented in Figure 6.1. Supplementary information to 
stereogrammes can also include a clear description of the 
characteristics of the objects and statements describing the 
significance of each type. 

Descriptive characteristics of vegetation corresponding to image 
aspects relevant to photo analysis are the texture (related to forest 
cover density) ; the height of trees and the size of dominant tree 
crowns; and the tonal variation of the different vegetation storeys. 
A ground description of accessibility such as water courses, and the 
impact of man's .action on the environment, can also be a valuable 
information to be included with stereogrammes. The use of 
stereogrammes ensures a good control of photo -interpret at ion and their 
application results in a more uniform and consistent stratification. 

Forest stand delineation 

This delineation is concerned with the further division of forest 
types into mangrove forest stands. These units, which are often used 
as compartment units, consists of even-aged or otherwise homogeneous 
stands which are to be subjected to identical silvicultural and other 
treatments. A forest stand is thus an operational unit, which is 
identified during the forest management planning process and used 
extensively in the operational management of any forest. 

If recent aerial photos on a large scale, say 1:10 000 are 
available, and the area has bean under intensive management using the 
clear-felling system for a long time, it will be quite easy to 
distinguish the. different stands on these photos. However, this is 
rarely the case when it comes to mangroves, and it is thus necessary to 
combine these with extensive surveys in the field. 


Ml : Low mangrove foreet. Partially marahy with email but *ene tree crowns 
with a few large tree*. Avicennia sp. dominant. 

M2 : Medium high mangrove forest with a mixture of Ml and N3 . 

M3 : High mangrove for* it with large tree crowni. 

Pm : Mareh land, ealty water, occurring between mangrove etandt and lower area*. 

Fig. 6.1: Stereogramme showing manfrove forest types 

Courtesy: Sicco-Smit 


6.2.3 Operational level surveys 

On this level the surveys most often consist of either field 
enumerations (and/or observations) or area surveying on the ground, to 
complement the surveys mentioned above and to record changes since the 
last survey was undertaken. 

Field enumeration to gather information on wood resources are 
described in the next chapter and a few case studies are presented in 
the back of this manual. 

Land or area surveying is described in numerous forestry and 
surveying manuals and will therefore not be dealt with in this manual. 
Due to the constraints in conducting such surveys in the mangrove areas 
(the occurrence of many rivers and streams, the sometimes very soft 
ground, the high tides which make the recording of land near the 
borders and rivers rather difficult and the often very high erosion and 
accretion rates) , it is recommended only to undertake such surveys of 
smaller areas and to obtain recent aerial photos in a sufficiently 
large scale to aid in any mapping and area estimation. 


The presentation of survey results is often done in the form of a map. 
The purpose of making a map can be for the planning of more detailed surveys, 
or to assist in decision making regarding the use and development of resources 
at different application levels. With particular reference to mangrove 
forests, the information required for each application includes various items. 
A few of them are presented in Table 6.3 below. 

Table 6J: Type of Information and map scale requirements for different application levels 


Type of 

Scale range 

National level 

Planning level 

Planning level 

-Geographical distribution 
of mangroves - area extent 
-Broad lind use 

-Mangrove sites 
-Broad forest classes 

-Mangrove forest 
-Forest stands and types 

1:50 000 to 1:250 000 

1:25 000 to 1:50 000 

1:25 000 

The accuracy and amount of the plotted detail should be correlated with 
the potential value of the land it represents. Also, in more intensively 
managed forests, scales should be larger as more data must be reported on a 
smaller area. 

At a national level, mapping consists of presenting the general 
distribution of mangroves of a country or a region. Orbiting satellite 
imagery, combined with small to medium scale aerial photography are suitable 
for such a task. 

The information which can be put on the map may include the following: 

Forest lands (natural and planted) ; 

Lands used for aquaculture (fish ponds/shrimp farms); 

Agricultural lands within mangrove areas; 

Mining and industrial zones; 

Infrastructure, settlements and urban areas. 

Figure 6.2 is an illustration of a small scale map showing the extent 
of mangrove forest and other land uses. 

At the medium level (management), semi-detailed land-use mapping can be 
carried out with the objective of producing maps at medium scales, on which 
various forest sites can be shown. The latter can be defined by forest 
density and development conditions. They may be: 

Areas where mangrove forests are well preserved and can be allocated to 
timber production and some kind of intensive forest management could be 
imposed, and 

Areas destined for conservation and protection purposes or allocated to 
other utilizations than timber production due to the nature of the 
forest stand. 

Maps at medium scale can also give information on forest land and land- 
use distribution. This is illustrated in Figure 6.3. 

Intensive forest management applications call for detailed stand type 
maps with a good planimetric acuracy. Forest stand classification should 
provide up-to-date information on certain important parameters including tree 
species or species groups, age classes, regeneration, cutting activities, 
degree v of stocking etc. The classification of forest lands into distinct 
forest types can only be achieved with aerial photography having sufficiently 
high resolution, complemented with ground observations. Figure 6.4 is an 
illustration of a managed mangrove forest, where compartments are shown and 
progress of cutting is indicated. For more detailed information on map and 
mosaic compilation please refer to Appendix 3. 


Different techniques have been used to develop area estimates. The most 
common is the use of a planimeter on a pre-established planimetric map of the 
entire area. Measurements are made on maps which contain photographic details 
such as forest types. Area measurements can also be performed directly on 
aerial photographs of relatively flat terrain, another common advantage in 
mangrove areas. 

The main drawbacks of this method is however the shape of the mangrove 
areas, which are so often intercepted by rivers and creeks, and the size and 
shape of the minimum area in the classification units. Figures 5.8, 5.10 end 
6.2 are examples illustrating this point. 


Fl|. 6 J: 

A small scale land-use map baaed on digital Landsat image classification with a 

Scale: 1:230000 

(Source: Tikumponvarokisrta/., 1985) 

Tropical tvtrgnen forist 
HMjrovi 4ornt 
Rubbir plintitiMi 
Deteriorated fornti 

I I Matr and unclassified 

Grasiland/Savannah/paddy fields 


Fig. 6.3: A medium scale land-use and forest type map based on a SPOT image 

(Same area as in Fig. 5.6) 
(Ngomeni, Kenya) 
(Source: Lantieri, 1986) 

17*773 ZA: Agricultural zone 

C: Non agricultural zone 
A: Village 


Inter -tidal zone limit 

T8: Dry barren landa 

TH: Wet barren land* 

TVl: Barren land with low vegetation cover 

TV2: Barren land with high vegetation cover 

BZ: Flooded ponda . 

B8: Dry ponda 

EL: Mater 

S3: Dry coral aanda 

SH: Wet coral aandt 

N: Rhizophora 

A: Avicennia 

Dl: 30% of trees with no leavea 

D2: 100% of tree* with no leavea 

D3: Cut areas. Wood on ground 

D4: Non -mangrove wood. Wet soil 


Fig . 6.4: A large scale mangrove management map showing compartments and logging areas 

(Matang Mangrove Forest Reserve, Malaysia) 
(Courtesy: Forest Dept, Malaysia) 


Several sampling procedures including dot grid and transect methods have 
also been applied. Among these, the dot grid procedure is considered the most 
advantageous in practice. 

The technique is easy to use and is widely applied. It consists of 
laying a transparent grid with equally spaced dots upon a map or a photograph. 
Each dot falling within a given class or a stratum is counted and used to 
estimate the total area of the class. The technique is described in detail 
in several forest inventory books and documents (Spurr, 1952; Loetsch and 
Haller, 1973) . 

If K is the total number of dots falling on the area of study and Kj is 
the total number of dots counted in stratum j, the area proportion of stratum 
j is obtained by: 


Pj = -- 

When the dot grid is used with aerial photography, the scale variation 
can cause the area estimates to be biased, but on level terrain such as in 
mangroves, scale variation is minor and dot counts can be used without 

The dot grid and other sampling techniques are also affected by the 
minimum area on which photo- interpretation is based. The size of the smallest 
area unit varies according to photo scale and survey intensity. 

Forestry literature shows that minimum areas between 1 ha and 20 ha have 
been commonly used. Moesner (1957) found that area proportion estimates and 
survey costs increase with increasing minimum area size.. The cost is 
certainly affected by the size of the minimum area, but mostly by the increase 
in variability of vegetation, making photo- interpretation more difficult. For 
area determination however, it is generally recommended to use small minimum 
areas, because area estimates are more accurate. 

The most important limitation of the dot grid procedure is the 
determination of the error on area estimates. The statistical problem 
encountered with the dot grid technique is that dots are systematically 
distributed over the area, and the utilization of the binomial model assuming 
a random distribution is not adequate. In order to get around this conceptual 
problem, several error formulas with varying degree of complexity, have been 

Chevrou (1979) suggests the following easy to use and correct expression: 

CV (S)% - 50 P' 5 N 7fc 

CV is the coefficient of variation or the relative error 
S is the area estimate 

N is the number of dots falling in the area S 
P is the perimeter ratio i.e. the perimeter of S divided by 
the perimeter of a circle of the same size. 

When using image data or maps in digital form it is often possible to 
obtain an area estimate directly without any manual measurements apart from 
the ones needed to check the accuracy of the map and its scale in the field. 




Forest resource assessment is here used in the sense of a resource 
appraisal, a rough estimate of the wood resources available. This estimate 
is neither as accurate nor as detailed as the results from a forest inventory, 
but the method is found useful, when a rapid assessment of the resources is 
needed . 

This assessment can be based on either measurements taken directly on 
remote sensing imagery or on limited field sampling or, of course, on a 
combination of the two techniques. 

7.1.1 Volume estimation from remote sensing imagery 

The rising costs of conventional inventory techniques and the 
urgent need for information have led forest mensurationists and 
managers to use procedures of volume estimation from remote sensing 
imagery in particularly aerial photography. Several forest and tree 
characteristics can be measured on aerial photographs of adequate scale 
and quality, and aerial photography has been used for direct volume 
estimation for many years. 

The general procedure in volume estimation of a tree or a stand 
is the establishment of a mathematical model where the independent 
variables can be measured or estimated directly from aerial photos. 
These components of photo volume are related to the volume measured on 
the ground by regression techniques. 

Height measurements 

Among the parameters used, tree height has been found to be the 
variable most closely related to volume. The parallax method using 
simple devices, such as parallax wedge or parallax bar, is most 
commonly applied. Height accuracy has been tested in several studies 
and it is known to be dependent to a large extend on the flying height 
and camera type. Best results can be obtained with very large scale 
imagery, but fairly good height assessment could be achieved on medium 
scale (1:10 000 - 1:20 000) as well. The formula below is used for 
height determination: 

D P *H 
h - 

b + D p 


h is the tree or stand height 

H is the flying height above ground 

D p is the parallax difference between the top and the foot 

of the tree 
b is the average photo air base 


In order to increase the accuracy of the flying height, and 
subsequently of tree height measurements and photo scale, efficient and 
practical instruments have been developed. The foliage penetrating 
radar altimeter is an example. It can be attached to photographic 
systems on board a light aircraft to obtain reliable distance from the 
plane to the ground. This informatibn is most of the times lacking 
when a light aircraft is used for photo coverage with small camera 

Tree crown measurements 

Tree crown width, crown closure and crown area are other 
parameters which are also introduced in volume equation computations. 
They are directly measured on aerial photographs by means of simple 
devices. While crown area and crown width are determined for single 
tree volume equations, crown closure, which measures the percentage of 
crown cover, is used for stand volume equation determination. Crown 
density scales are constructed with black dots on white background or 
vice- versa. The density of dots represents different ranges of crown 
density. Crown density is determined by comparison of the standard of 
the scale and the photo density on plots. 

Stand density 

In some situations, stand density proved to be an interesting 
parameter to combine with crown closure for volume determination. Tree 
counts, however, become increasingly inaccurate as the photo scale 
becomes smaller, especially in young stands. 

Volume estimation 

Numerous studies have been undertaken to establish the 
relationship between the above parameters and the actual volume on the 
ground for various tree species and forest types found in the inland 
forests and savannas. However, only few studies of this kind have been 
carried out in mangrove areas. Khan, Choudhury and Islam (1990) report 
on attempts to study the possibility of estimating stand volume in the 
Sundarbans using variables which can be measured on aerial photographs. 
Fourteen stand volume equations were solved using combined photo and 
ground variables as independent variables. The standard error and 
correlation coefficient were estimated and the equations tested against 
corresponding volume calculated from ground data. Whereas the use of 
the basal area as the independent variable gave the best results, and 
the standard error of the estimate was 16% when using both photo and 
ground variables, it was found that four different equations using only 
variables, which can be obtained from aerial photos (number of 
trees/ha, crown cover percentage and average stand height) gave a 
standard error of the estimate in the range of 18-19% with correlation 
co-efficients of 0.887-0.898. These equations are presented in 
Appendix 5. However, it should be noted that in the above experiments 
the photo variables were in fact obtained from the field survey and the 
results thus only give an indication of the possibilities of using 
measurements taken directly on aerial photographs of a suitable scale 
(1:10 000 or 1:5 000 for timber volume estimation). 


The same authors also reported on attempts to estimate the stand 
volume using Landsat data, but none of the initial equations showed 
satisfactory results. 

7.1.2 Volume estimation with limited field sampling 

Using this method, measurements of the diameter and height of the 
trees in a few sampling plots representative of the forest types (or 
other units of classification) are undertaken in the field. With the 
help of pre-established volume tables the volume in each plot is 
calculated, and the average volume for each of the forest 
types/classification units is then multiplied by the total area of each 
type and added up to obtain an estimate of the total volume of wood 
available. For more details refer to Section 5.2 in this chapter. 

The sampling intensity is often very low, and the sampling units 
neither randomly nor evenly distributed over the entire area, but often 
located in clusters in easily accessible areas. The volume t'ables used 
are likewise not verified and/or revised to reflect the local 

The estimate obtained is thus not very accurate and it is 
impossible to determine the error on the result. As very little 
information is available on the mangrove resources, this method is 
however valuable in obtaining a first indication of the extent of the 
resources on a national/regional scale. 


Forest inventories are concerned with more detailed and accurate 
estimates of especially the standing volume of wood. As such, they provide 
valuable information needed in the preparation of forest management plans and 
in preparation and execution of operational plans such as logging plans, where 
what is generally desired is a very detailed knowledge about the quantity and 
quality of the wood available and a reliable estimate of the size of the area 
where logging operations will take place. 

This information requires generally an intensive inventory, mostly in 
the field, but maps and aerial photography are valuable* aids in the 
preparation of the inventory in the location of compartments and tracts 
concerned by it. 

7.2.1 Sampling designs 

Strip and line plot sampling techniques are commonly applied in 
tropical forests. In the following sections, a brief account is given 
as to their application in mangrove forests. Other sampling designs 
are outlined in Appendix 4. 

Sampling lines oriented perpendicular to the waterways often 
provide good data capture due to the natural tendency for mangroves to 
exhibit zonations parallel to the waterways. The difficult working 
environment and general absence of landmarks also favour systematic 
line sampling procedures. 

Strip sampling 

This sampling design consists of the laying out of continuous 
strips of uniform width, running across the topographic gradient and 
drainage pattern so as to cover most stand conditions. The strips can 
be randomly selected but in practice, regularly spaced strips are 
generally used. Figure 7.1 is an illustration of a strip layout. 

The sample strips are normally laid out at distances between 500 
and 1 000 m and have a width of 10 to 20 m, depending on the sampling 

In practice, strip sampling poses difficulties in the field, 
where a dense understorey and windfalls are frequent. In mangrove 
areas this is an even more acute problem because the progression in the 
field often is hindered by a muddy ground and dense stilt-roots, and 
the ensuing difficulty in maintaining a constant width of the strip 
results in considerable errors. 

In addition, for the same sampling intensity, the number of 
sampling units using the strip sampling method is relatively small, 
compared to the line plot sampling method, which from a statistical 
point of view is rather unfortunate. Another disadvantage of strip 
sampling is the fact that its long narrow shape has a long edge, 
relative to its area, resulting in frequent border tree occurrence. 

When strip sampling is used, field observations can be recorded 
using a cumulative tally for all strips, or a separate tally by strip 
or even within strips. The data is recorded in individual segments 
which are then the sampling units. 

The strips may be equal or unequal in length. When they are 
unequal, regression or ratio estimator procedures should be used for 
estimating the mean volume and standard error of the mean. This 
provision also holds true for the case of strips divided into segments 
which in turn may be unequal in size, particularly at the ends of 

An example of strip sampling is presented in Gas* Study 2. 

Line plot sampling 

In line plot sampling, a set of field plots, generally of the 
same size, is located throughout the area of interest. For practical 
reasons, a systematic layout of plots (see Figure 7.2) is used in most 
inventories, where this technique is employed, in spite of the 
statistical difficulty associated with the estimation of the variance 
of the estimate. 

For a given sampling intensity, different layout schemes can be 
devised, depending on the sample size, distance between plots on the 
line and distance between lines. In Thailand for instance, the line 
plot adopted consists of lines 100 m apart with circular plots of 5.64 
m radius (100 m 2 ) 100 m apart on the line. This scheme gives a 1% 
sampling intensity. 


fig. 7.1: An illustration of a systematic strip layout. 

Flf. 7.2: An UhKtration of a line plot layout. 


However, a more typical distribution of plots is where the 
distance between lines is greater than the distance between plots on a 
line. In this way, fewer lines are used, but because more plots occur 
on the same line, it is more likely that most vegetation and stand 
conditions are represented in the sample. Because of the zonal 
distribution of mangrove vegetation, this is generally ensured when 
lines are oriented in a direction perpendicular to topographic gradient 
and drainage pattern. This configuration is for instance adopted in 
the mangrove forest inventories of Matang, Malaysia, where the distance 
between plots on the line is 20m and the lines are 100 m apart. Each 
circular plot has an area of about 79 m a (5m radius) resulting in a 4% 
sampling intensity. Please also refer to Case Study 4. 

When choosing a sampling intensity for a systematic line plot 
sampling, it might be useful to carry out a few trials to determine the 
variability of stand parameters. See also Section 7.2.2. 

A line plot design can be drawn up as follows: 

Let A be the total area concerned by the inventory, 
Ap the forest area which is actually tallied, 

a the area of the sampling unit, 

n the number of sampling units, 

f the sampling intensity, 
Dl the distance between lines and 
Dp the distance between plots on the line, 



Ap m f*A ; n ; 


To determine plot or line intervals, either Dl or Dp has to be 
chosen on the basis of practical convenience. 

When the sampling units are distributed according to a grid 
system, the calculation of the mean, the total and their variance is 
often carried out as though the units were randomly chosen. A closer 
approximation of the true variance of the mean can be obtained if the 
sum of squares of differences between successive plots is used instead 
of squares of deviations from the mean as in simple random sampling. 
Such an approximation is given by the following formula: 

L n, 

[x,, - x 
2 jl i-1 
S_ - (1-f) 

x L 

2n E (n r l) 



S_ - Standard error of the mean 


L * number of lines 

n 3 m number of sampling units in line j 

n = total number of units in the sample 

f * n/N with N being the total number of sampling 
units in the population 

Compared with the strip method, the line plot design has the 
following advantages: 

The delineation of strips is not necessary, therefore the 
progress in the forest is easier and less cumbersome. 

The uniform distribution of plots over the whole area provides 
more accurate data, especially in mangrove areas where the forest 
conditions vary from deep flooding to dry ground zones. 

7.2.2 Sampling intensity 

When a sampling inventory is to be implemented, the intensity of 
the sampling should be defined. It is a function of, among other 
things, the forest extent and accessibility, the sampling design 
chosen, the funds available and the precision required. Techniques of 
enumeration and sampling intensity will differ according to whether: 

a) an extensive resource assessment is planned, the results of which 
may be required for large administrative units, or 

b) the results of the inventory are to be used for intensive 
management, timber sale and exploitation which require fairly 
high accurate information on small area units or stands. 

In general, a compromise between the cost of enumeration and the 
precision of the results has to be found. Obviously, the optimum 
solution is to achieve a high precision with the least cost, but in 
addition to all these theoretical considerations, feasibility and 
practicality are important factors. 

In the case of a simple random sampling for which the precision 
has been specified, the number of sampling units n is given by the 
general formula: 


n * 

t a s a 

E> + 


t denotes the value of Student's t corresponding to a 

probability level (1-a) 

s 2 is an estimate of the population variance 
N is the total number of units in the whole population 
B is the maximum allowable error on the estimate. 


The above mentioned formula can take on other forms and can be 
simplified if N is large (infinite population) to the following: 

t's 2 

E 2 

The cost of the inventory may also be fixed and the inventory is 
to be executed within a restricted allotment of funds. Considering the 
total cost C to be: 

C = CO + nCl 

Where CO is the overhead cost and Cl is the cost of enumerating 
one sampling unit. The number of sampling units can then be computed 

C - CO 


In such a case, the precision of estimation is computed according 
to the resources available. 

Where mangrove forests are characterized by relatively 
homogeneous stands and the species are of moderate economic value, a 
sampling intensity of about 2% is generally considered to be 

7.2.3 Sampling unit shape and size 

The effect of plot size and shape on the variance of the 
estimates has been the subject of numerous studies. With regard to 
plot shape, the main factor influencing the choice should be the length 
of the perimeter and the ease of plot establishment. Circular plots 
should theoretically be more efficient as they have the largest 
area/perimeter ratio, minimizing therefore the occurrence of borderline 
trees. The walking time during plot measurement is also minimized for 
circular plots. 

However, while some authors recommend the use of circular plots, 
others have shown that long and narrow plots can be efficient along a 
topographic or a fertility gradient. In practice, circular plots are 
often satisfactory and are extensively used in forest inventories, but 
for research purposes, such as permanent plots used in continuous 
forest inventories, square and rectangular plots may be more 
appropriate, as they are easier to demarcate and relocate accurately. 

The sampling unit size is another item to determine before the 
survey starts. Theoretically, since the precision of estimates depends 
on the number of units - for the same intensity of sampling - units of 
small areas are more efficient than large area units. Conversely, 
several studies have shown that to some extend, large size units are 
associated with a small coefficient of variation, which is the primary 
variable that affects the sample size. 


With regard to mangrove areas, small plots are usually adopted. 
This is because mangrove stands often are very homogeneous. Also by 
using small plots, the chance of having more than one forest type in 
the plot is small. 

In mangrove forests, this takes on a particular importance 
because, in addition to the possibility of introducing bias, the 
procedure of moving the plot so as to include it entirely in one 
stratum for instance, may turn out to be more laborious in the field 
than to enumerate all portions of the plot which fall in different 

In any case, it might be necessary in the specific mangrove 
conditions and local situation, to determine the sampling unit size on 
the basis of preliminary trials using various kinds of sampling units, 
and choose the most adequate size on the basis of maximum precision, 
cost and practical convenience, keeping in mind that th optimum six* 
of tha sampling units is an arta which contains an average of 10 to 25 
trtts to b* measured. 

In natural and/or mature mangrove stands, bigger plot sizes might 
thus be needed. See the discussion in Case Study 4. 

7.2.4 Continuous forest inventory (CFI) 

The actual trend in forest inventories for management purposes is 
to use information based on data from permanent plots. It may also be 
useful to consider combining permanent sampling units in the field and 
on aerial photography. This approach is an efficient method for 
assessing changes (Schmid-Haas, 1980) . 

The use of permanent plot based information for management 
decisions, requires that their choice and establishment be judicious. 
They must be representative of the varying forest conditions, and be 
subjected to the same silvicultural treatments as the non-sampled part 
of the forest. 

To fulfill the requirement of their establishment, aerial 
photography is an invaluable aid for their location and also for their 
periodic relocation. It must be stressed that one should give a precise 
description of the sites, using bearing and distance measurements, 
topographic details and land marks. 

To ensure that permanent plots will be subjected to similar 
treatments as the remaining parts of the stand where they are located, 
it is recommended to use concealed marks. Instead of tagging 
individual trees in the plot for instance, stem locations may be mapped 
on a plot diagram sheet, with their coordinate position and number. 

Since the data collected on permanent plots will be used in the 
determination of yield and growth and the evaluation of changes over 
time, care should also be taken during field enumeration to ensure 
consistency with regard to the measurements of tree characteristics. 

Plot information that may be recorded in each sampling unit 
includes the following items: 


Plot data Individual traa data 

Date of measurement Tree Number 

Plot number and location Species 

Forest type Diameter (DBH and other) 

Stand size and condition Total height 

Density (class) Form and quality 

Soil type Vigor 

Inundation class 

Understorey vegetation 

7.3.1 Tree characteristic measurements 

As in other forest inventories, the diameter and the height are 
the main variables to be measured. The principle of their measurements 
and their relationships with volume are discussed in most inventory 
books. In the following discussion, it is however attempted to present 
a few practical aspects relevant to their measurements in the 
particular mangrove conditions. 

Diameter (DBH) 

The diameter can be measured with a calliper or a tape. The 
problem of error in dbh measurement, caused by eccentricity or oval 
form at breast height is of minor importance for most mangrove species, 
whose stem form is relatively normal, and a tape is thus often used as 
it is less cumbersome. In the highly moist conditions and salty air in 
mangroves, fiber glass tapes are preferred to steel tapes. 

More important than the choice of measuring tool, is the level at 
which measurements should be taken. The general rules used in other 
forest formations are applicable. Some modifications may be necessary 
according to the local conditions and practical convenience. An 
important exception however, concerns the mangrove trees with stilt- 
roots, such as Rhizophora sp., where the diameter measurements should 
be taken at 30 cm above the highest root resulting in 'a deformation of 
the stem (FAO, 1981) . This might in some cases be as much as 4-6 m 
above ground level. 

As to the diameter anomalies that may be encountered at the point 
of measurement, such as a fork, a swelling, or other abnormal form, 
conventional measurements procedures may be applied. 


Height is often measured with a clinometer. In mangrove forests, 
instruments which do not require distance measurements may be 
preferable. A cheap and handy instrument of this kind is the Christen 
hypsometer. Optical range finder may also be useful for a rapid height 
evaluation. The observer should however be placed directly under the 
tree to be measured to avoid over estimation of tree height. 


In low stands, height measurements can be obtained with a 
telescopic rod, A wooden or bamboo stick - bearing a graduation 
scale - can also be used. 

Because of the importance of height /diameter functions in 
providing information on the structural conditions of the stands, a few 
studies have been conducted in mangrove forest and several models have 
been established, (see Sandrasegaran, 1971) . 


The bark of some mangrove species is commercially valuable. Bark 
measurements must therefore be taken to evaluate the production. For 
timber estimation, the conversion of timber over bark to timber without 
bark (merchantable timber) requires the knowledge of the bark volume or 
proportion. Bark thickness measurements are made by means of a bark 
gauge or simply with a ruler after the bark has been stripped off on a 
limited area of the stem. 

7.3.2 Volume determination 

On standing trees the volume is usually determined indirectly 
through volume tables or volume functions. Generally, the volume is 
expressed as a function of dbh or girth sometimes combined with the 
height. In most mangrove forests where volume functions or tables have 
been established in former days, the data were not always adequate. 
During more recent forest inventories however, equations based on 
sufficient observations have been determined for some areas and 
species. (See for instance Boonyobhas, 1986 and ODA, 1985 as well as 
the examples given in Apondix 5) . 

Volume tables are usually monospecific but, in tropical forests 
with mixed species having the same utilization and growing in the same 
conditions, volume equations may be developed for mixed species stands. 
It is worth noting that in mangrove forests of relatively homogeneous 
stands, it might be useful to consider simplified volume functions 
based on the mean tree. The total stocking can be determined by 
multiplying the volume of the mean tree (the tree with the mean basal 
area) by the number of stems per unit area. 

On felled trees volume determination involves the application of 
the universal log formulas (known as Huber, Smalian and Newton 
formulas) . Refer to Cast Study 5 for a description of the steps 
involved in the construction of a local volume table for Rhizophora 

Determination of bark volume 

Volume estimates often include the bark, which in some cases is 
a valuable product in itself, and measurements should therefore be 
taken in order to assess the volume or weight of the bark and to revise 
the volume estimates accordingly. 

Table 7.1 on the following page shows an analysis of bark volume 
for Rhizophora mangle and R. harrisonii selectively felled and measured 
at Playa Garza In Costa Rica (Chong, 1988) . 


table 7.1: 

Diuutcr b.h. (e) 

Penxnt of total bwk 

for BMsophoro numgU/*. herriaonii trM 









Total Volume 









2 . 0504 


Total Volum 






. 9647 





Bark volum* 



















The bark-factor method (Meyer's method) may also be used to 
determine bark volume. An example is given in Box 7.1 on the following 

7.3.3 Growth determination 

Attempts have been made to assess dbh, height and volume growth 
of mangrove species in several countries. However, in mangrove 
forests, yield and increment figures have in most cases so far been 
based on little and fragmentary data. 

Whereas some attempts have been made recently to measure the 
diameter increment of mangroves through studies of growth rings in 
countries with a distinct annual dry season, the assessment of growth 
of mangrove trees can best be determined from permanent plots by means 
of periodic measurements of trees and stands, and growth determination 
is thus a lengthy and difficult endeavor. However, growth date are 
essential for forest management planning, in the determination of the 
forest yield potential and for establishing the optimal silvicultural 
system to be applied. 


The presentation of resource assessment results is often given in the 
form of a simple table showing the size of the area covered by each of the 
different forest types - corresponding to the type map produced earlier; the 
estimate of average standing volume /ha for each type and the total volume for 
the forest area in question. 

Forest inventory results are more detailed and often relate to a smaller 
area such as a forest district. A compartment map is often available and some 
of the inventory results may be incorporated in the compartment register. 
However, separate tables and histogrammes are often used for presenting the 
bulk of information obtained. 

For each major species or forest type, the area of each ageclass (or 
diameter class, when age is not known) is calculated and the result presented 
in a tabular form and/or illustrated in a histogramme . The more even the 
distribution, the closer the forest is to the 'normal forest', and the easier 
it is to maintain a constant yield over time and thus ensure sustainable 
forest management (See Chapter 9 for further information on the concept of 
the 'normal forest'). 


In the Trr*ba-Sifp reserve fe Costa Rica, *$? flfiftaafiaa fi^HZfltf trees 
were felled and rtieaaured to establish the relatioriship between diameters overbark Id ) 
and underbade <d B ) at fertast height. A plot of d u as a function of d e gave a ttnear 
relationship with a y intercept ctose to the origin (0). The predictive equation for this 
rnaytnua be written In the generat form d - fcd , where kis the regression 

coef ctent or tort factor. As the regression coefficient k (0.93d) wa$ derived from 
diameter measurements taken at breast height, it is referred teas the to western bark 

Given n average value for k, the bait volume (V b ) for any log section may be 

Baric volume 



Volume overbark - Volume underbark 
tV B ) 

diameter overbark at mid-section; 
diameter underbark at mid-section; 
sectional length; 
volume overbark 


Tharafore V fc 


V, - V, * V. - k*(V,). 

BafkVetuma V*f%J * 


tn practice, V fc derived by equation (C) wll be greater than the actual 'stacked 
volume' because overbark atawietar meauremerrt uauaHy made with diameter tapes 
include the air spaces between the ridges of the bark whereas the stacked volume is 
smaller because of bark compression. The stacked volume for unpeeled toga will be 
amafler than that calculated In equation (C) and a correctkm factor may be included to 
equation C as for example: 

whare a correction factor of 0.8 was used. 

Box 7.1: DetenatoaiioB of barit 


Stand tables for each major species/forest type showing the standing 
volume for each age/diameter class are also produced. 

An example of a stand table for Rhizonhora. harriaonil/R. man ale and 
Pelliclera rhizovhorae based on an inventory of 63 plots in Playa Garza, Costa 
Rica is shown in Table 7.2 Note however, that this table shows number of 
trees/ha for each diameter class rather than standing volume/ha. 

Table 7.2: Stand table for Rhiwphoras/PeWciera rhiwphorae 









































412 1 










769 1 

Source: Chong, (1988a) 

An illustration of a stand table for a 'normal forest 1 
Figure 7.3 below. A uniform annual increment is assumed. 

is shown in 

0-9 10-19 20-29 

Age doss (ywrs) 


Figure 7.3: An Illustration of a stand table for a 'noimalfonsf 

Local volume tables may be presented either in a tabular form with one 
or two entries or expressed as a volume regression and presented graphically. 
Examples are given in Caw Study 5 ind Appendix AS* 



This part focuses on the application of biological, managerial, 
technical, economic and social knowledge, and manpower resources to manage the 
use of mangrove resources in a way that will provide sustainable benefits to 
the greatest number of people without impairing the environment. 



Integrated management planning presupposes that the greatest societal 
benefits are realised when forests are managed for a mix of goods and services 
on a sustainable basis* 

Wood and non-wood potential uses and their sustainable economic 
implications are analysed. A multiple-use strategy that harmonizes viable 
uses is formulated. Unlike traditional planning approaches, timber production 
is not over -emphasized at the expense of non-wood components of the ecosystem. 
Nowhere is the need to strike a balance between different uses more compelling 
than in mangroves, where more often than not, the non-wood opportunities may 
be economically and socially more important. Planning, therefore, is required 
to achieve the desired combination of forest uses over space and time, so that 
the various productive components of the forest production system can be 
optimally used and sustained to meet intended objectives. 

Ideally, forest management should be based on a complete understanding 
of all of the social, economic and ecological parameters that are involved. 
That the current knowledge on the ecological functioning and biological 
interactions for some non-wood resources is incomplete, is however, not in 
itself an impediment to their sustainable management. Forest management is 
enhanced through .practice, and its scientific foundation is strengthened 
through trials and experimentation. Management may be conducted empirically 
on the basis of limited inventory supplemented by eperience, sound reasoning 
and intuition. 

A key question for the longer term is how to reconcile sound 
environmental activities with economic growth expectations. Part of the 
answer may well be founded on the design of policies that will foster a 
pattern of economic growth that makes use of a wider range and integrated mix 
of resource opportunities rather than over-exploiting any single-use. This 
diversified and less single resource -intensive approach is one aspect of 
integrated multiple-use planning. 

Mangrove management planning should thus be part of an Integrated 
Coastal Area Management (ICAM) programme. I CAM ensures the sustainable use 
of the economic goods and services generated by the coastal ecosystems for 
meeting development objectives and to preserve the environmental health, 
resource quality and ecological integrity of the coastal area. 


On a practical basis, as non-wood uses are usually managed by non- 
forestry agencies, it follows that coordination and linkages between concerned 
land uses and the relevant agencies /users will be required. Many different 
uses of the mangrove area (such as for instance production of wood, bee 
keeping, coastal protection and small scale capture fisheries) are totally 
compatible and can be carried out simultaneously. Others (such as large scale 
aquaculture, protection of wildlife habitats and intensive forest operations) 
are less so and a zonation of the area according to priority uses might be 

Whereas integrated management of mangroves is strongly advocated, the 
focus of the present document is on the forest management aspects and 
additional information on management of other mangrove resources such as 
fisheries and wildlife must be sought elsewhere. 


Although it is essential that mangrove resources be .managed on a 
sustainable basis, a purely biological approach towards resource management 
is often unacceptable because the needs of socieMw may be quite different and 
very often do not conform to the capability of the land to support those 
needs. For example, in South Vietnam it was necessary to allot some prime 
mangroves for shrimp- farming because the local people could not survive on 
forestry activities alone. In other less populated areas, the forests may be 
able to produce more than is needed by the local communities and transport 
costs may prohibit the export of wood to other areas. 

To formulate appropriate forest use plans, the peoples' demand for goods 
and services should therefore first be determined. The demand may be local 
or regional. In the economic sense, demand is gauged in terms of the cost, 
quality and location of the service to be provided in relation to alternate 
and substitute supplies. As forest benefits are not infinite, the use of one 
form of resource will often be at the expense of other alternative uses. For 
this reason, the loss of other opportunities for using a mix of resources must 
also be considered. 

The next step is to analyse to which extent these demands can be met 
from the mangrove area in question, based on the assessment of current and 
potential resources. 


"The Jbasic fault in the conventional approach is that the rural poor are 
rarely consulted in planning or given an active role in development 
activities. This is because the poor have no organizational structure to 
represent their interests 91 . The first task therefore is to assess the needs 
of the direct and indirect beneficiaries and direct planning towards meeting 
as much as possible the needs of the target groups. The lesson is clear: 
unless the rural poor are given the means to participate fully in development, 
they will be excluded from its benefits. 


Planning for a multiple-use management is a complex task, in that 
problems must be viewed from different perspectives and needs. It will be 
necessary to relate the policy to be applied at regional, divisional and local 
levels to reduce planning conflicts. 

Similarly, long term and short-term goals at different levels must also 
be harmonized. In most mangrove areas, a significant part of the production 
may be used to meet demands outside the forest area. The Ayeyarwady mangroves 
in Myanmar for instance, is highly depleted and degraded due to the high urban 
charcoal demand in Yangon estimated at some 700 000 t/year. Until recently 
about *500 000 tonnes were annually supplied from the mangroves. In other 
cases, fuelwood prices become so high that they exceed the buying propensity 
of local villagers, forcing them to cut fuelwood from public forests. 
Generally speaking, forest plans should be flexible enough to accommodate 
changes in political, economic and environmental conditions. 

The forest plans should furthermore be part of an ICAM programme that 
is built on a multisectoral framework, which seeks to harmonize environmental 
linkages by minimizing the spillover effects of various types of sectorial 
development and by re-allocating and sometimes reducing access to natural 
resources . 


The following principles are used as a guide in forest planning: 

(a) Wood and non-wood resources are managed and used to meet local, 
regional or national needs: 

The importance of a resource supply is not determined by it's 
physical or biological characteristics but by the priority that society 
places on its use. Development of national and regional policies will 
establish the required emphasis between the various sectors at a cost 
that is appropriate for the total level of returns desired. Planning 
will shape policies into programmes that are compatible with local 

(b) An assessment of needs and public participation is an integral part of 
the planning process: 

Managing natural resources to meet peoples' needs implies a 
knowledge of what they want. The people may hold views which are 
coloured by local customs, religious or other values. The involvement 
of people in the planning process is used as a tool for gaining 
information about peoples' views, values and priorities. 

(c) Plans must be obi active oriented: 

When the problems or issues are understood, a set of objectives 
should be framed to address key issues. Objectives should be 
quantifiable targets that serve to focus management effort and measure 
performance . 


(d) Plans must trv to achieve the greatest good for the greatest number of 
people in the long run! 

Minority interests must be weighed in relation to the general 
well-being of larger communties. In practice it is impossible to 
achieve a complete or unanimous support -for all planning decisions. 

(e) The ecological carrying capacity should never be exceeded and resource 
sust a inability should be given high priority! 

This is a non- negotiable requirement, if sustainable development 
is to be achieved. This requirement should be given high priority in 
the planning agenda. 

(f) The need for biodiversity and wildlife conservation should be 

This should be incorporated into the plan appropriate to the 
scale of the management area. For a small and/or highly fragmented 
area, it will be impractical to reserve large tracts of pristine 
vegetation for conservation purposes. Instead, the establishment of 
well placed control plots may be more feasible. 

(g) Planning is an on-going dynamic process: 

Planning must be flexible enough to accommodate shifts in 
demand/supplies and priorities. Because societal values change over 
time, planning is an on-going dynamic process. Change must be 

Generally the larger the geographic unit, the longer the planning 
horizon (time- frame) . Regional policy objectives are necessarily long- 
term and are based on general trends that are affected only by macro 
changes. Forest District Management Plans, on the other hand, are 
based on medium- term plan objectives and are revised more frequently as 
the information base expands. Operational plans are short- term. (See 
Table 8.1) . 

(h) The plan must provide for improvements in data collection to reduce 
areas of uncertainty associated with an incomplete or weak information 

The ultimate objective may be achieved in phases, taking into 
account an improved information base over time and applying a 

conservative approach where the uncertainty is perceived to be great. 

(i) The decision-making process must be visible and equitable: 

Involving the public in the decision-making process is necessary 
to promote local support and acceptance for integrated forest 
management planning. Just as it will be the duty of the forest service 
to explain to the public the implications of various decisions, the 
greatest value from the public will most likely be in using their 
knowledge of local conditions and needs. 


Customary rights should be respected where possible. Decision- 
making should not marginalize the traditional incomes of local people 
nor their access to reasonable amount of forest products without 
offering practical and acceptable alternatives. 

(j) Planning functions and responsibilities 

The responsibility for planning functions should be clearly spelt 
out at different levels. Typically a national Forest Service is 
headed by a Director General of Forestry (Chief Conservator) , who is 
assisted by his Deputy DCs (Deputy Chief Conservators/Assistant Chief 
Conservators) , supported by several territorially based State/ 
Divisional/Provincial Forest Officers (Conservators/Regional Forest 
Officers) and District Forest Officers (Deputy /Assist ant Conservators) . 

The terms used within brackets will, undoubtedly, be familiar to 
those who have worked under the British colonial forestry services. 
The Forest Service as a department is under the direction of a Minister 
in charge of forestry matters. 

Planning may be undertaken by a macro-micro planning cell or by 
the Working Plans Branch or Division within the Forest Service. The 
different forest management levels and responsibilities are shown in 
Table 8.1 as a guide. 


i* . 


H J3 

> 4J 


a 8df 

iH ***. 

rH 4J -H tt 


JH tn 4J fc 

a 8 -S-3* 

ua fl 3-3 

38 82 




H 4) 
h <M 

* 2*8 

S 6 ^? 





i JJ O 



O 4J -H kl 
-H 01 4J -H 

b I llj 

.H -2 "2 h - 

rt AJ TJ 

C -H JJ ) 4J 

-2 S 85 8 

xi -H b w ^ 


H O 

S u g 
H 8 8 

o " $ 

o JQ o 






S-U *W -H 9 
-H -H E 4J 

- b ti 5 fe 

p, fi I) O 
Vi TJ I) Q 

8-. g 




b AJ 

-S S 


-H rt 


? ! 1 - 5 

r-i o tn 




CO V4 -H 


I 5 

a -a 2 


V XJ -H tH rl 

H W OJ *M 3 rH S 

tfc, *O 

-3^ - S 


o t) a tn 



" 8 B 

H b 

H -H 


b O 

O M-i 



|| o 






I. 8 

4 4 

Oi o 9 i- 

Si HI 



Mangrove forest management is based on the sciences and skills of 
geology, pedology, climatology, hydrology, botany, ecology, silviculture, 
forest technology and economics in the selection and treatment of both wood 
and non-wood resources. 

A concise plan, setting out the requirements and controls to be applied 
and the activities to be implemented over space and time in a logical 
sequence, to achieve desired objectives is referred to as a management plan. 

Such a plan can be a resource and development document applicable to a 
country or a region; a forest aanagtatnt plan/working plan for a forest 
reserve or forest district, or an operational plan for (part of) a forest 
tract . 


The basic planning steps applicable to each planning level, with minor 
modifications, may be described as follows: 

(a) Setting the Terms of Reference 

Define the management area, the planning horizon, the financial 
and human resources and the time-frame allotted to undertake the tasks. 
This will not be a problem in a plan revision exercise, where the area 
is known and past survey cost data is available. For an unmanaged 
area, however, attention should be focused on what is practicable and 
affordable . 

(b) Assemble baseline information 

Relevant socioeconomic, ecological and resource data are 
collected, compiled and documented in a structured format. Existing 
maps, available data and past inventory records are consulted and 

(c) Identify constraints 

Constraints are generally inflexible but may be circumvented in 
some cases. For example, if the tract of forests to be managed is too 
small, a switch to higher value-added products or service management 
may justify the operating cost involved. Alternatively, where land is 
available the forest estate could be enlarged through land acquisition 
or reservation. Constraints are categorized as follows: 

(1) Technical /biological ! Technical or biological factors may 
constrain the extraction methods to be applied or the products to 
be produced. For example, site limitations will restrict the 
species that can be established. 

(2) Financial : The rate of return on capital may be insufficient to 
meet the rigid standards set by lending institutions. 


.(3) Socioeconomic! A plan cannot not operate in isolation. The 
resources allotted to its use will become unavailable for other 
uses. The overall benefit to the community involves employment 
generated, environmental impact, and "invisible 11 benefits derived 
from savings in other sectors, such as improved fishery, 
ecotourism, coastal protection, etc. Local customs, culture and 
religious beliefs may constrain the use and promotion of certain 
products or services, such as alcohol from fermented Nvpa sap or 
wild boar meat. 

(4) Institutional: These are limitations imposed by the organization 
and managerial ability of the body executing the plan, the legal 
framework, social pattern and attitudes, low literacy rates, etc. 

(d) Formulate objectives 

Production goals should be designed to meet as much of the 
societal needs for each resource use as possible within the limits of 
sustainability. Other goals regarding the environment, soils and water 
protection and rural development are also considered. 

(e) Develop management alternatives 

Where economic and financial data are available, management 
alternatives may be compared in terms of their cost -effectiveness, 
taking into account other equally valid considerations, viz. social, 
cultural and environmental factors. The choice and ranking of 
priorities will depend on the alternatives that can best achieve the 
preferred set of objectives. 

(f ) Prepare Management plan 

"Management plan" is here used in the generic sense to include 
plans applicable to each planning level. As .mentioned earlier, this 
plan should be part of an Integrated Coastal Area Management programme 
to ensure sustainable multiple use of the mangrove resources. 

(g) Implementing the plan 

An activity schedule to implement plan targets is drawn up. 
Further data may need to be collected, as for example regeneration 
sampling prior to logging. 

(h) Monitor and evaluate plan results 

Periodic review of plan outputs is required to see how well 
objectives are being met and to make adjustments as required. To 
facilitate the evaluation process, "criteria" or "indicators" for 
measuring the success or efficiency of the adopted plan are drawn up. 
A desirable criterion should, (1) provide in a single figure all the 
information needed to make the decision; (2) be applicable to all 
alternatives, and (3) be readily calculable. 

The Government's criteria might be: 

number of jobs created and their location; 

income generation and their distributional effects; 

impact on foreign exchange; 

the IRR, NDR, and/or Cost/Benefit ratio. 

The Forest Department's criteria might be: 

rate of tree growth and areas of plantation created; 

profitability (revenue/expenditure relationship) ; 

rate of deforestation/degradation. 

The lending institutions' criteria might include: 

rate of loan repayment; 

total capital sum invested. 

From the conservation perspective/ if we accept the premise that 
biodiversity is the variety, number of different species, and the 
quantity of each species in a forest and its associated environment, 
then the criteria for acceptance of a chosen management alternative 
must adequately reflect these concerns. This may give rise to 
conflicting interests. In any given location, the number of species 
will change due to ecological succession and longevity of species. 
However, a management system which maintains trees as the key 
structural element in the ecological landscape has a better chance of 
maintaining biodiversity than a degraded environment. 


Different management planning levels are tied to geographical units. 
(Table 8.1) . Thm regional oanagtMnt plan is territorially divided into a 
number of forests or forest districts which are self sustainable units. It 
should be noted that/ whereas provinces and districts are civil administrative 
units, which may be demarcated socio-politically, forest districts are 
delineated according to natural terrain features, which may or may not 
coincide with the administrative units above. As the area covered is 
extensive, the planning horizon is necessarily long-term, because large 
investments are needed for plan implementation. Regional forest plans often 
have a timeframe of 10-20 years. 

At the forest management /working plan level, the management area is most 
likely to be a forest district, often constituted as a Forest Reserve. Tha 
fortftt m*n*g*m*nt plan covers all of the forest and although predicted 
removals and a felling plan are prepared for the whole rotation (say 25-30 
years) , the plan period may be ten years or less due to the difficulties of 
forecasting the economic as well as the demand situation over long periods. 

Th* working plan on the other hand only covers areas in which forest 
operations are to be undertaken within the working plan period, which is often 
shorter than the forest management plan for the district in order to take 
account of new factors or changes (normally 5-10 years) . The working plan may 
be further divided into separate plans covering silviculture/ harvesting 
operation etc. 


An operational plan entails a further division of the area in that it 
deals with detailed specifications for on-site operations to be carried out 
in the near future (1-3 years at most) and may be prepared for each range 
within the forest. 

9.3.1 Data types 

Data needs should be clarified before embarking on data 
collection in order to save time and money. Data are collected to 
assist in formulating realistic courses of action, to allow the 
possible courses of action to be evaluated and thus ultimately to 
facilitate the decision-making process. Five classes of data are 
required as follows: 

(a) resource data; 

(b) operational data; 

(c) utilisation data; 

(d) socioeconomic data; 

(e) institutional data. 

9.3.2 Resource data 

For each resource, the main information required is (a) 
availability, (b) productivity, and (c) cost. The relevant data are 
summarized as follows: 

Land area and type 

Tree cover, fisheries, agriculture 

Material and equipment 


Human resources 

For the assessment of the land area and the mangrove resources, 
please refer to Part III of these guidelines. The productivity of the 
mangrove forest resources is described in Chapter 11. The remaining 
resource data are collected as for other forest types and thus not 
presented in detail in this document. 

9.3.3 Operational data 

The preferred operational method should be prescribed and the 
work activities defined as follows: (a) extent covered (ha or km); (b) 
the input (man-days, machine hours, materials, etc.); (c) the output 
(ha/day, km/day); and, (d) the cost per unit area or effort. The 
anticipated increase in MAI or survival rates are useful benchmarks for 
measuring performance. The operational data requirement may usefully 
be summarized as follows: 

Land clearing - including timber harvesting, hauling, burning, 

pond construction, canal construction, etc.: 

Site preparation * for afforestation/reforestation; 

Nursery establishment, propagule collection, etc; 


Maintenance and protection; 

Improvement and production control; 

River transportation logistics, canals, etc. 

Please refer to Chapters 10-12 for details. 
9.3.4 Utilisation data 

There should be an effective demand- forecast ing system for the 
forest products and service expected to be produced from the forests at 
various levels. Thus this data type is required for timber/fuelwood 
producing mangroves. Even for environmental management areas, the 
objective may change over time to include production functions. 
Relevant factors to consider are as follows: 

Preferred species - type and properties; 

Spacing - size assortment, log size, quality, etc.; 

Area - wood volume, size and types of kilns, other cottage industries; 

Growth rate -production schedule; 

Location of forest, processing units and transport facilities: 

Site conditions affecting logging costs; 

Profile on traditional wood/nonwood uses 

Moat of these data can be collected from a review of past management and 
utilization and/or from specialized demand studies. 


9.3.5 Socioeconomic data 

Economic considerations are required over and above purely 
financial ones simply because strict analysis of cash expenditure and 
revenue do not fully account for the real cost and the real benefit to 
the community as a whole. In multiple-use management timber production 
may be reduced or even curtailed to preserve or enhance aquatic 
production and the trade-offs between alternatives compared. As the 
economic quantification of intangible costs is still at an early stage 
as applied to forestry in general and mangroves in particular, and 
informed estimates may be used instead. 

The socioeconomic data needed are as follows: 

Shadow labour costs; 

Labour opportunity costs; 

Associated social costs - e.g. public investments in housing, water supply, 

canalization, crossings, etc; 

Discount rate to be used; 

Shadow price for produce to reflect price distortions due to taxes, duties 

and price control mechanisms; 

Value of non-marketable benefits - e.g. improved environment, health, 

shelter, erosion control, recreation, etc.; 

Development impact of intangible benefits to local or regional economy, 

training, etc. 

9.3.6 Institutional 

Institutional factors are mainly political by nature, but also 
include the legal framework. The management plan should include the 
following statements: 

Legal obligations; 

Legal privileges and rights; 

Policy guidelines; 

Support to communities, education and training; 

Reactional facilities; 

Local attitudes and impact on local society; 

Research linkage and support. 


The first task is to define the productive forest potential, and 
identify the main constraints. [6.1 (c)] The productive, protective, and 
social functions of the forests are defined as the objects of management, 
according to priorities. Whereas goals refer to the desired long-term 
perspective, objectives refer to measurable activities (outputs) prescribed 
within the plan period. Examples of management objectives are given in Box 
9.1. Sustained production of wood and non-wood forest produce is an important 
function. Protective functions, inter alia, include the following: 

Riverine and coastal protection; 

conserving wild plants and animals through habitat management; 

Preserving unique forest stands or ecosystem; 

: Production forestry 

(1) To produce a sustained yield of quality greenwood for charcoal processing to meet 
domestic demand as well as for export; 

(2) To produce bark (tannin) as a by-product of (1); 

(3) To produce quality poles, posts for local consumption; 

(4) To produce a sustained yield of firewood to supplement die domestic and industrial 
fceJwood and energy requirement* of ftie nation; 

(5) To produce fishing stakes, and structural materials for the local communities; 

(6) To produce such other related mangrove forest products that may be required for 
tertiary or rural cottage industries; 

(7) To plan for integrated utilization of mangrove resources. 

(1) To protect, rehabilitate and manage mangrove ecosystems that are required as 
breeding ground or source of nutrition or shelter for shrimps, molluscs, fish and other 
high protein sea-foods; 

(2) To maintain die integrity of mangrove vegetation along the coasts and estuaries to 
erve as storm barriers, flood and erosion control; and to provide environmental 
support and protection to coastal agricultural cropping and communities; 

<3) To preserve and keep inviolate sufficient areas of natural mangrove ecosystems as 
reservoirs of specie* diversity and for conservation of plant and animal genetic 

(4) To set aside sufficient areas as may be required for research, education and training 

(5) To manage area* required for recreation and/or tourism. 

(6) To promote social acceptance for forestry, better utilization and forest management. 

(7) To generally regulate the we of waterways, channels and creeks within the 
mangroves so that tbeir navigational value wtfl not be impaired. 

Box 9.1 Management objectives 

In Vietnam, mangrove forest enterprises are managed as profit centres. 
Funds generated from the forests are used to finance forest development and 
to improve the social wellbeing of communities in the enterprise area. 


The production strategy may comprise a mix of products ranging from 
fish, shrimps, charcoal, poles, timber, Nvna thatch, bricks, etc. (Refer to 
Box 9.2 below) 








Generally to manage, develop and protect the mangrove resources in order to achieve 
producttoti of wood and non-wood benefits in ottler lo ftilfWocaJ and coastal demaoto 
construction materials and other wood product*; 

Spedficolty to produce a sustained yield of die following products for local consumption at 
affordable and stable prices: 

quality greenwood for charcoal processing fat domestic cooking industrial and after 

energy requirements; 

firewood at equitable prices; 

fishing stakes, poles, posts, and structural materials for local coronwmtict; 

adequate sawlogs for local use; 

bark (tannin-based dyes) 

related mangrove forest products, including nipa, for tertiary <* cottage processing 


Generally to improve the standard of living and quality of life of the mangrove dependent 
population (including fishermen and shrimp-farmers) within fte ttai Mai forest enterprise with 
particular regard to the following;- 

sustainable meaningful employment; 

adequate and improved quality of bousing; 

availability of essential medical care; 

adequate and Improved education facilities; 

improved cultural and community facilities; 

adequate and affordable communication services to promote greater social interwakm, 

marketing of products and client distribution of social goods and services; and, 

enhance the scenic and amenity values of Ac forest, 

Specifically through extension, demonstration and training activities increase the peoples 1 
awareness and acceptance of forestry programmes. 

Generatfy to maintain the integrity of mangrove vegetation along the cowtt and rivers totem as 
storm barriers, Hood and erosion control; and to provide environmental support and protection to 
coastal agriculture, aquaculture tod homesteads; ; 

Specifically to protect, rehabilitate and manage mangrove ecosystems Oat are required as 
breeding ground or source of nutrition or shelter for shrimpy molluscs, fish and other W^protw 

preserve natural mangrove ecosystems as tesovote of species diversity and for 

conservation of plant and amraal genetic resources; 

set aside sufficient areas rcqeired forresetrcb, educatiofi tftd ttitmng ptiipow; 

manage recreation and tourism areas; 

promote social accepttuwe to tow 

maintain navigability of channels and waterways. 

Box 9.2: Dat Mui Forest Enterprise management goals and objectives 



Based on the objectives of management, a plan strategy is chosen to 
reflect the local conditions. In countries such as Vietnam, where local 
authorities are non-existent or are unable to provide the basic social 
amenities to the mangrove dependent communities, it will for example often be 
the responsibility of the local forest enterprise to take on rural community 
development activities as well. During the plan period the strategy may thus 
include, inter alia, some of measures listed in Box 9.3 below. 




promote sound local land use management and coordinate with other land 

upgrade the technical, managerial, ami extension capability of the 

tbe tewaew knowledge of the fore* enterprise through 
<Hveraficat*on of investments to increase its profitability so as to provide a 
wider range of social benefit* to target beneficiaries; 
promote the welfare of forest workers by providing adequate 
health, recreation, education and training facilities; 

devise a resettlement programme for toe mangrove dependent population 

otto less ecologically fragile and agriculturally suitable sites; 

emphasize on people^riemed management and participation through 


develop an adequate local energy plan; 

apply rational toultiple-use forest management. 

develop an gfapaie and sufficient energy plan; 

ftifler utilization and minimize processing wastes; 

promote rural cottage industries aid otter rural income generating 

opportunities and 

Source: Cbong, (I949a) 

Box 9.3: Plan strategy for * I brat enterprise in Vietnam 

Another example of a plan strategy is from the Sierpe-Terraba mangroves in 
Costa Rica, where, given the large extent of irregular forests that were 
selectively logged and the need to regularize and introduce management control 
as quickly as possible, the strategy outlined in Box 9.4 was recommended to 
augment and conserve the fuelwood resource. 

Whenever possible, alternative management strategies should be drawn up 
and evaluated according to their ability to meet the desired set of management 
objectives with due regard to the concept of sustainable development of the 
mangrove forest ecosystem. 



and rett ovw-cta stoi^te; " , Cf 

regulate and Improve removal; '',,T 

silvJcutturat trawform fowt into unrto^ 
create local village woodtot; 

incorporate environmental monitoring and control measure*; * * 
upgrade the mangrove management capability of the Forest Service; and, 
provide tocaJ toriaiog ^ 


, (188) 

Box 9.4: Plan strategy for the Sferpe-Terraba mangroves in Costa Rica 


The concept of sustainable development with regard to forest resources 
has often been described as the sustained yield principle. This term, 
however, brings connotations of the production of timber and does not evoke 
the production of environmental services, which - especially in the case of 
mangroves - are of equal importance. A holistic approach is thus needed and 
the environmental impacts of different management strategies, including the 
production of wood, should be evaluated in order to ensure the sustainable 
development of the entire ecosystem. 

With regard to wood production, sustainable use implies that the yield 
of wood products should never fall. In mangrove areas (as in other forests) 
which have previously not been subjected to management regimes, this concept 
is too rigid, as the forest may be very heterogenous or contain an abundance 
of overmature trees and thus far from the concept of a 'normal forest 1 often 
used as a goal to aim for. See Box 9.5 below. 

A normal forest is an ideally constituted forest with such volumes of trees of 
various ages so distributed and growing in such a way that they produce equ 
volume of produce, which can be removed continuously without detriment to 
future productton OTd erwironment. 

Box 9.5: DeHiilUonofa^rmalforcst* 

This ideal norm is seldom achieved in all parts of the forest due to 
ecological changes, varying market demands and other unforeseen factors such 
as pests and diseases. 

In areas where wood production is the primary objective, a conversion 
phase will normally be required to transform irregular forests into managed, 
even-aged forests of higher productivity. During this phase, an idea of what 
is biologically, technically and economically feasible in restructuring of the 
growing stock is crystallized. The lack of management planning information 
on growth dynamics and other aspects, as well as the need to introduce forest 
management as soon as possible, may make it expedient to draw experiences from 
other sources or countries. 


Tentative standards for achieving required final crop stocking are 
applied, and these are refined as the informational database improves. 
However, blanket prescriptions for all forest types are not advisable because 
mangroves are dynamic ecosystems and in spite of their broad structural 
similarities are nevertheless very site-specific. 

To achieve sustainable development in the long term, some immature 
stands may have to be felled during the conversion phase so that age-classes 
are brought closer to the normal distribution, implying that sacrifices have 
to be made in order to achieve sustainable production in the future. Only the 
management planner can advise whether the sacrifice is too great or not, 
bearing in mind that unwanted trees today may become commercial lumber 
tomorrow . 

As normality may not be achievable within a single rotation, annual 
removals should be kept reasonably flexible. Depending on demand, some 
over-cutting is permissible, provided periodic adjustments are made so that 
removals do not exceed the biological potential of the forest. Whereas 
individual stands may be cut too early or too late and the annual production 
may vary, a sustained production should be maintained within a working circle 
during the management plan period. 


The regional aanagenent plan is territorially divided into a number of 
forests or forest districts, which are self sustainable units under the 
responsibility of a District Forest Officer. 

Forests are often legally constituted as Forest Reserves, which are 
dedicated to forest management for wood production and non-wood benefits. 
Reserves are subdivided into a number of compartments . A 'compartment 7 is the 
smallest administrative unit of management, location and record that is 
territorial and permanently defined for purposes of description and record. 
They are demarcated on the ground according to natural boundaries and defined 
by surveyed maps. The compartment size varies according to the intensity of 
work to be undertaken. The first step in organising forest management is to 
prepare a compartment map and a schedule of works for the survey, demarcation 
and identification of compartments on the ground. Only forest reserves which 
will be worked during the current plan period need to be demarcated as this 
is a costly and time-consuming activity. 

A sub- compartment is a treatment unit. It may be defined as a 
subdivision of a compartment, generally temporary in nature, differentiated 
for special description and treatment. Sub- compartments function as 
silvicultural and production operational units, whereas compartments have 
administrative, managerial functions. 

The forest estate is also silvicultural organized into a number of units 
as follows: 

(a) Working circle (W.C.)? Sub-compartments are grouped under different 
'working circles' (W.C J . A W.C. is defined as 'an area (forming the 
whole or part of a working plan area) organized with a particular 
object and under one silvicultural system and one set of working plan 
prescriptions' . 


Allotment of sub- compartments to W.C.s depends on site factors, forest 
types and silvicultural treatment needed according to objectives df 
management. Overlapping W.C.s may be necessary when, for example, the 
same compartments are endowed not only with molluscan resources but 
also productive stands requiring different types of treatment* 
Alternatively, W.C.s may be designated for recreation and wildlife or 
for coastal protection purposes. 

For control and supervision purposes, W.C.s are subdivided into 
territorial ranges under the control of a range forester (forest 
supervisor) . In large ranges, there are several beats, each of which 
is supervised by a forest guard. 

(b) Felling Series: To administer harvesting and regeneration operations, 
and to provide stable employment, a working circle may be divided into 
two or more felling series. A 'felling series' is an area of forest 
delimited for management purposes and forming either the whole or part 
of a working circle. Its aims are, firstly to create a distribution of 
felling and regeneration areas attuned to local conditions, and 
secondly, to maintain or promote *a balanced age-class distribution. 
Each felling series is a sustained yield unit. 

(c) The Periodic Block; Felling operations may be organized according to 
periodic blocks for more or less even-aged forests. A Periodic block 
is defined as the part or parts of a forest set aside to be regenerated 
or otherwise treated during a specified period. Regeneration may be 
secured through one or more regeneration fellings or through artificial 
regeneration. If the rotation (R) is 25 years and the regeneration 
period (RP) is 5 years, the number of periodic blocks is five, with 
each block having one-fifth of the felling series. The area of the 
periodic block (A) for a felling series (FS) of 1 000 ha will be 200 
ha. The formula to derive the periodic block area is A FS x RP/R. 

Periodic blocks may be "permanent, revocable or single". In the first 
case the blocks are permanently selected and cannot be changed; in the 
revocable method stands may be reassigned to other blocks, and in the 
last method priority is always given to blocks based silvicultural 
regeneration, utilization and site factors. The last method may be 
considered to be a special case of the revocable periodic block method. 


The 'working plan' is the part of the Forest Management Plan which deals 
with the prescriptions of the work to be undertaken within the plan period and 
as an operational blue print comprises the silviculture and treatment plan, 
the felling plan, revenue collection arrangements and financial forecast, a 
basic description of the working plan area and its parts including a map 
together with guidelines and priorities for area management. 

A working plan must contain a clear statement showing the average annual 
cut in for a fixed duration, say the next 10 years. Clear prescriptions are 
given to achieve the wood production targets and forest regeneration 
programmes and other silvicultural treatments planned. The major components 
of a working plan are discussed in the following sections. 



A system for managing the production side of forestry must be in place 
to support intended management objectives and operational goals. Producing 
goods and services on a continuing basis confers many social and economic 
advantages, which are beneficial to rural communities particularly for 
sustainable fuelwood supplies. A silvicultural plan is a means to gradually 
transform the forest stands into more manageable and efficient productive 
entities. An exception to this, are forests earmarked for preservation or for 
conversion to other non-wood uses. 


"A silvicultural system may be defined as the process by which the crops 
constituting a forest are tended, removed, and replaced by new crops, 
resulting in the production of stands of distinctive form" (Matthews, 1989) . 
The silvicultural system to be applied depends on the ease with which the 
desiraJble species can regenerate themselves naturally in the disturbed 
environment caused by logging and/or the degree to which they lend themselves 
to artificial regeneration methods. There are few ready made systems which 
can be used directly without some adaptations to suit local situations. A 
silvicultural system comprises 3 main components as follows: 

(1) the method of regeneration chosen suited to local ecologies, site 
potential and silvics of preferred species; 

(2) the form of crop produced; and 

(3) the systematic arrangement of the crops over the whole forest 
estate, with reference to silvicultural and protective 
considerations and efficient harvesting of produce. 

A classification of silvicultural systems is shown in Box 10.1. 

Some silviculture systems that may be applied to mangroves are briefly 
discussed and their main advantages and disadvantages summarized in the 

10.1.1 Clear-felling systems 

Clear-felling systems aim to establish an even-aged stand by 
removing the mature stands in a single operation. Where the principal 
species are light demanding and can regenerate naturally, and the sites 
are favourable, such systems may be very cost-effective. The Matang 
mangroves have been managed over three rotations using clear-felling 
systems in blocks without any major problems, except that large areas 
have to be artificially regenerated as more marginal sites are brought 
under intensive management, with R. apiculata as the main species. 

The visual impact after logging can be very disconcerting to non- 
foresters and conservationists. It should not be practised in areas 
where ecotourism is contemplated and the felling coupes should not be 
too extensive. The pros and cons of such systems are listed in 
Box 10.2. 


Al HIGH FOREST SYSTEMS jm^l&t&tog^''^ 

81 ' Pcfliag/regcaentioo confined to only fwt. of foiwti " ; . ' ; /; ^& ; ' : 

Cl: Crop ckued by k WHag, resulting in even-cd ; cwp; 

' ' 

C2: Crop cleared by succesrivt regenewton foUiags; mwtoBg 
inmons orlwa cv^Mtgod cnqpt; 

Dl: Regeneraiioti <listributtd conjptrtmea-wisc 

a. Canopy ojxming oven; young CJDJM more or less 

cvtMged tnd unifonn crop: U*tf*m Systems 

b. Canopy qpeatag is scattemi gapt, ymu^ cropt 

mm or toss ^ven tged ingipt; <^N(Sp Sij^tow 

c. Canopy uprabig irregular and gradual, rwotoog 

cropiomcwbat uneven^god: Irrtgvbr Shtterw**! 

D2: Regeoention 


s. Pelting in strips: 

b, Felling in wedge pattern: 

B2 Felling/regeneration distributed continuously over (be whcde area; crop 
completely irregular or uaeven-aged: S&ctio* Systems 

Vtriaflt systems produced by: 

a, introducing & young crop beneath an existing 

immature one: Two-storM * fl%* Fam* 

b, retaining certain old crop trees after regeneration 

is completed: High Far&t wMh Jtoem 

A2 COPPICE SYSTEMS (mainly vegetative propagation) 

1. Stands derived entirely from vegetative shoots: 

Crop even-aged tod clear felted: Cc^pfee System 

Crop uneven-aged and partially feBed: ftypfcr* JMMb S^t 

2. Stands comprise partly vegeoaive shoots and partly from trees 

of seedling origin: Ctpptce irj* StasAmb 

A3 SHELTERWOOD SYSTEMS a generic tenn describing systems ofwcessive 
ftto&tt aad selertions systems, 

Atof^; *\wi~*g*d* ^ synonymous with "uniform* or *r9$ul*r** Tfie t*m* 
byJ.D. M*tih*wa, 

Box 10.1: Sllvfcuttural systems 

A "clear-felling in alternate strips" system, with and without 
retained standards (seed-bearers) is practised in several countries 
(Thailand, Venezuela, Cuba and Costa Rica) . Felling strips are 
aesthetically more acceptable. The prospects of natural regeneration 
are enhanced due to the narrow width and the long borders relative to 
the size of the area felled and management control is simple to apply. 


(t) riapte tfr HBJ^^ 

(^) lof^g cam tie geoenBy lower, higher 

(c) over-BMtoe ittiri removed in u* operttioa; 

$ kss ddfli needdd ttafl otter j*prodm*^ *aethodi; 

(e) create evHi$ed regulated ftatm taw* rotation; 

$ afford* complete overhead tight, required by fight demaaders, 

(a) erosion ad she deterioration rah my be higher; 

(b) twdlfaagi miy be wewdy dtoibuted; 

(c) ipemi man be *bte to 

(d) ti&4m*& totdvme* pawtto tf togging IK< wfl cooduaed; 

^) i^toM* tibe awtbctic Md fflttMrity vtiuw Ofstancb; 

(f) tlltitca, iiro^xoive ofspeciti iotf merchtatobility ait cut; 

(g) koitttto a bfge amooat of togging abit tad 

Box it J: Advai^agestnddisadvai^agesof Clear-feUlnjSyst^ 

This system, due to its simplicity, is recommended where there is 
a shortage of trained personnel and/or skilled workers. It is also 
suitable for those countries where mangroves are newly brought under 

10.1.2 Selection systems 

Selection systems are characterised by two conditions: viz., the 
stands are uneven-aged, and the forest cover is never completely 
removed so as to deprive advance growth and seedlings of shelter and 
shade. Generally, such systems favour shade tolerant species but the 
degree of canopy opening may be manipulated to favour light demanders 
as well (e.g. Group Selection) 

A selection system has been practised in the Sundarbans for a 
long time and also in the Ayeyarwady mangroves in Myanmar. This is an 
environment- friendly system in that the merchantable trees are 
harvested periodically and over all parts of the forests. In practice, 
however, unless the forests are adequately stocked, and the technical 
and subordinate staff are well trained coupled with responsible timber 
contractors, management can be very complex. The merits and demerits 
of selection systems are summarized in Box 10.3. 

A variant of the Selection method is Group Selection. This 
system creates larger felling gaps, that favours the regeneration of 
light demanding species and promotes the formation of small groups of 
even-aged stands. Consequently, harvesting costs are lower and wood 
extraction la simpler. 


(ft) only system capable of maintaining aa uaeveiHged 

(b) reproduction of tokMat tpeciei is easfy obtained; 

(c) ike protection fe excellent; 

(d) stands era be readily adapted to meet ft&tuating n 
() capital returns *t short totcrvab. 


level of tedmkal stilb m& nraigewat cottrnl needwl; 
extraction costs are hi^ier and iflttlfcr iBov*li/uift *M; 
product dimemiom acre vwwble; 
crop tree* we saoeped throughout the ttmfc 
inventory dau analyi tad grow&-yield fomnaats e 4ffiBndt 
not favouribte for growing iatofenat spectra 

Box 10 J: Advantafles and disadvantages of Selection System 

Sheherwood Systems 


provides protection to species with sensitive juvenile stage; 

excellent soil protection and reduce* invasion by weeds; 

less risk of multiplication of injurious insects that breed in clearing!; 

stands more wind-film and belter adapted to cyclonic areas; 

aesthetically more pleasing man ctear-Wling systems; 

selected trees can put on better increments (hrough stand improvement 

treatments and gap openings. 

requires more skill; 

work dispersed, felting and extraction ten profitable; 

higher logging damage to young crops; 

delayed regeneration response can be cosily; 

yield regulation and silviculture mace complex. 

Box 10.4: 

d^dvairtaiesrf Sheherwood Syrtons 


In the Sierpe-Terraba mangroves in Costa Rica, a Selection system 
was suggested for areas with proven molluscan potential, as the 
commercial bivalve Piangua (Anadara tuber culosa, Sowerby) , which 
appears to be associated with Rhizophora and Pelliciera roots, does not 
thrive under open conditions unlike their Asian varieties. Partial 
removal of the overwood reduces site disturbance and over-exposure. 

10.1.3 Shelter wood systems 

Shelterwood systems are those high forest systems in which the 
young crop is established under the shade or side- shelter of the old 
one, referred to as the "overwood". The overwood protects the site and 
sustains the forest micro -environment conducive to the regeneration and 
growth of the younger trees. The term is used to include some variants 
of the selection system using successive regeneration fellings. Pros 
and cons of such systems are described in Box 10.4 on the previous 


A species preference list, based on silvicultural and marketing 
requirements, should be drawn up as a guide in prioritising treatment. The 
species, which are selected as 
deilrablef, vary according to 
ecosystem type, location and market 
demand. In Costa 
harri soni i and R . 
desirables , whereas 


3. Pelliciera rhizophorae 

4. Avicennia germ/nans 

Rica, Rhizophora 

mangle are the 

in Sierra Leone, 

West Africa, R. mangle is a dwarf 
form and R. racemosa is the preferred 
species. In Malaysia, Thailand and 
the Mekong Delta in Vietnam 
Rhizophora conjugata, R. apiculata 
and R. mucronata are highly favoured. 
In the Bangladeshi Sundarbans, Sundri 
(Heritiera fames) is the prime timber 
species, followed by Gewa (Excoecaria 
agallocha) a proven pulp species. In the Guanal mangroves in Cuba Avicennia 
yerminans (mangle prieto) is the favoured species as the wood is suitable for 
railway ties and utility timber. An example of a species list for the Sierpe- 
Terraba mangroves, Costa Rica is shown in the box to the right. 


/. Rhizophora harrisonii 
2. Rhizophora mangled 

Acceptable 5. Laguncularia racemosa 
Undesirable 6. Other species 

The dwarf-form of Rhizophora mangle, 
probably an ecotype should be avoided. 


Advocates of natural regeneration argue that such silvicultural systems 
are more in tune with the natural indigenous forest ecologies. The pros and 
cons of natural versus artificial regeneration are described in Box 10.5 on 
the following page. 


L draper to establish; 

2. less labour and heavy equipment required; 

3. origin ofxl source* usually known; 

4. better early root development by natural seeflings; 

5. usually teas soil disturbance 

1. tan control over spacing, initial stocking and diftritatkm of seedlings; 

2. riskofieed tree tow; 

3. genetically improved stock not eaUy 

4. regener^ion delays and failures powibte; 

5. greater need for ncm-commercial tJiimung; 
6> stands not Hated to mechanised extraction; 

Box 10.5: Advantages and dfauhantages of Natural Regeneration 

10.3.1 Seed sources for natural regeneration 

In a "clear-felling in alternate strip" 
sources of regeneration are as follows: 

system, the natural 

(a) existing advance growth of seedlings/saplings, 

(b) seeding from perimeter trees around the felling strip, 

(c) seeding from standards (mother-trees), 

(d) water-borne propagules from adjacent stands, 

(e) propagules from felled trees 

10.3.2 Retention of standards (seed-bearers) 

For Rhizophora stands the number of standards (seed-bearers) 
required is about 12 trees/ha. These should be well distributed and 
strategically retained in areas with insufficient or no regeneration. 
Generally, more standards are needed towards the swamp interior because 
of its inherent lower regeneration potential. When logging coincides 
with a heavy seeding year, the number of standards may be reduced. 


The use of standards implies that the species can bear enough of 
propagules to reseed the site after logging and that the trees so 
retained can withstand exposure and isolation. In the drier north 
Pacific coast in Costa Rica, Jimenez has observed that only dominant 
stems of more than 15 cm DBH.ob bear propagules (per. com.). The 
number of propagules varies from 6-350 per tree. Rhizophor&a are prone 
to sun- scorching and medium- sized stems apparently can withstand 
over-exposure better. Windthrow is always a risk on sites with very 
soft soils or exposed to strong winds. 

Seed-bearers should be chosen and marked before logging. The 
criteria for the selection of standards or seed-bearers are as follows: 


medium-sized (>16sm DBH.ob), vigorous trees with healthy crowns, which 
are judged to be capable of surviving one rotation; 

trees bearing viable propagules or capable of producing viable propagules 
(healthy, unbroken crowns); 

avoid over-mature m very large dees because they; 

are y^ aad Ugbtniag; tad can cause substantial 

damage to rtgrowth when they feU; 

may not produce viable propagules; 

contain substantial wood volumes, which, if not removed, will 

significantly reduce economic yields; 

are prone to termite aaada aod tak scorching. 




trees may be selected iftbey are heatthy and bear viable 
propagules, tot should be avoided if possible; 

standards that are damaged during logging should be not be accepted; 

in fluid or unstable substrates more standards are selected and these 
should be kgroujM of twcnl^ 

only desirable species are chosen as standards, 

more standards should be selected and retained in the btckswamp treat, 

BoxlO.f: Criteria for the selection 


Artificial planting is required to restock blanks and sites with 
insufficient natural regeneration. However, natural regeneration is always 
preferred because it is cheaper. For this reason an assessment of the 
regeneration stocking before and after logging should be carried out. A 
linear regeneration sampling (LSM) will provide an overview of the site 
regeneration potential, in terms of seedling abundance, distribution, species 
and sizes. Some of the aspects to be considered are as follows: 


10.4.1 Inadequate regeneration 

Regeneration may be insufficient or absent due to the following: 

(a) incomplete removal of overwood (in the case of light demancters) ; 

(b) excessive damage during logging due to incomplete supervision 
and/or logging operation is too prolonged; 

(c) excessive amount of logging debris and not properly stacked; 

(d) unfavourable soil condition; 

(e) absence of standards, 

(f) excessive tidal wash (high energy sites) due to indiscriminate 
removal of the protective fringe trees; 

(g) weed competition (e.g.Acrostichum ferns). 

10.4.2 Regeneration classes 

Seedlings above 30 cm high are often referred to as "established 
regeneration", and those below are noted as "potential regeneration". 
Regeneration classes, recognized and recorded during LSM regeneration 
sampling, are shown in Box 10.7 below. 

However, it should be noted that this classification system 
should be adapted to local conditions as Ehizophora propagules for 
instance may well exceed 30 cm in length without being considered 
"established regeneration". 

Regeneration Description 
classes (RC) 

I Seedling* of over 30 cm but le$$ than U mm height; 

n Seedlings/sapiiags of 1*5 m or more w height but lew than 3 ra; 

in Sapling* of 3m or tiwre in height bm less thin 5cm 

Box 10.7: Regeneration classes 

10.4.3 Regeneration stocking adequacy standards 

For adequate natural regeneration a minimum of 2 500 well 
distributed seedlings per hectare (RC I), equivalent to a spacing of 4 
mVseedling is required for multiple-use JZhizophora plantations. For 
bioenergy plantations, based on short rotations, 10 000 - 20 000 
seedlings/ha may be required. The purpose here is to optimize above- 
ground biomass rather than to produce a mix of products like posts, 
poles and large size charcoal billets. 

Applying mortality rates of 50, 30 and 10 percent for the 1-10, 
11-20 and 21-30 year periods, the corresponding stand densities will be 
1,250, 875 and 788 trees/ha at the end of 10, 20 and 30 years (TibU 
10.1). The overall mortality rate applied is -3.78% per annum 


Table 141.1: Average stand density and mortality rate for Rhi&phora apiculata 


Mortality % 
per period 


2 500 - 1 250 

1 250 - 875 

875 - 788 

Artificial planting must be initiated if less than 70 percent of 

the stand is regenerated with the desired species within 3 years. In 

Acrostichum fern infested areas, adequate regeneration should be 
.secured at the end of two years. 

The regeneration adequacy stocking standards for linear 

regeneration sampling (LSM) with different plot sizes are shown in 
TabU 10.2 below. 

Table 10 J: Minimum seedling regeneration stocking for different sampling 
quadrat size (RC I) 

LSM Quadrat Area in Quadrats Min. stocking 

sizes hectare per ha. per quadrat 

2.5m x 2.5m 


1 600 


5 m x 5 m 




10m x 10m 




Note: seedlings: 30cm < "height <1.5m; rounded. 

10.4.4 Linear regeneration sampling 

Linear regeneration surveys based on systematically laid out 
lines are used to assess regeneration status. Parameters used in the 
analysis and interpretation of results include the following: 

(a) Stocking: This gives an indication of the completeness or 
distribution of the regeneration: 

(b) Abundance : This refers to the number of 
individuals /quadrat for stocked quadrats. It gives an 
indication of crowding for larger regeneration classes and 
regeneration potential in the case of seedling 

(c) Regeneration size; The "regeneration size classes" of the 
advance growth is recorded. (Refer to the regeneration 
classification described above) . 

10.4.5 Effective stocking 

To assess effective stocking the relative presence, abundance and 
sizes of all regeneration classes RC I-III are weighed. When larger 
sized seedlings/saplings (RC II/III) are present, their potential 
contribution to final crop stocking is considered, particularly when 
young seedlings (RC I) are found to be lacking or insufficient. The 
analysis and interpretation of regeneration sampling, therefore, will 
not be complete unless the role of larger regeneration classes is 

If the effective stocking is less than 70 percent, the cause 
should be determined by field inspection. If the void areas are 
plantable, artificial regeneration should be undertaken during the 
following planting season. All planting costs and silvicultural 
operations are recorded in the compartment history. On the other hand, 
if portions of the area are void because of deep flooding or other 
natural causes which make it impractical to replant, the affected 
portions (if large enough, say 2 ha) should be marked out in the 
compartment map and noted in the compartment history. 


In planting of mangroves, plants belonging to the Rhizophoraceae 
are the most common species used and the following description applies 
mainly to Rhizophora spp. For information on artificial regeneration 
of other species the reader is referred to Das and Siddiqi (1985) and 
Siddiqi et al. (1993) . 

10.5.1 Phenology 

The flowering and fruiting behaviour of the principal economic 
species should be studied to secure the timely collection of ripe 
propagules . 

The selection of plus trees as potential bearers, and the setting 
up of "seed orchards 11 may be advantageous where there is a shortage of 
quality seeds and/or a large scale plantation programme is planned. 

Rhizophoras produce propagules annually. In Costa Rica, R. 
harrisonii produces mature propagules mainly during June and July 
although there are some stragglers. R. mangle flowers more freely. 

In Sierra Leone, West Africa, the main fruiting season of 
Rhizophora racemoaa coincides with the beginning of the rainy season in 
May-JUly and the ripe propagules are easily recognized by the 
appearance of a 'collar' beneath the pericarp. 

In Malaysia, the principal Rhizophoras fruit during June to 
December. Preliminary studies indicate that most species flower and 
bear fruits several months earlier in the drier and stressed sites in 
the Ayeyarwady delta area in Myanmar. For instance, Aegiceras 
corniculatum flowers and fruits during May to mid-July on the drier 
sites, but flowers and bears fruits only during July and mid-August in 
the lower intertidal zone. This general trend applies to most species 
that naturally occur over a wide range of sites. 


Figure 10.1: Dense natural regeneration of Rhizophora spp., Matang, Malaysia 

Photo by M.L.Wilkie 

Figure 10 J: Artificial regeneration of Rhiwphora apiculota, Matang Malaysia 

Photo by M.L.Wilkie 


10.5.2 Collection of propacrules 

In Costa Rica, Rhizophora mangle and R. harrisonii propagules 
look rather similar, although the latter has a longer radical, more 
lenticellated and tapers more towards the shoot. Propagules are 
collected from mature R. harriaonii/R. mangle stands as there are 
varieties (ecotypes) that tend to produce mult i- stemmed or dwarfed 
forms. The "mangleros 11 can easily distinguish between mangle 
cabal lero (JR. harrisonii) and mangle gateador (R. mangle) , and they are 
contracted for seed collection. In the Matang, R. apiculata/R. 
mucronata propagules are collected by tender during June to December. 
Contractors deliver the seedlings to the planting sites and these are 
culled by the Forest Ranger. 

Two fruiting seasons are observed in the Ayeyarwady delta in 
Myanmar. In upper tidal zone ripe seeds are collected from mid-May to 
early June, before the monsoon starts. In the wetter zone, the 
collection season is from mid- July to early August. Ripe propagules 
are dark brown with a tinge of purple and are easily detached from the 
tree by shaking. Newly fallen propagules that float are also 
collected. As a guide, only ripe and healthy propagules of normal size 
and having well-formed radicals, that are unblemished by insect attack 
marks, are accepted. 

10.5.3 Site preparation 

After logging, the slash should be pruned, collected and stacked 
in neat rows perpendicular or 45 to the waterways. This is done to 
promote tidal flushing, dispersal of water-borne propagules and to 
reduce tidal ly induced slash movement that can cause damage to 
established seedlings and advance growth on the ground. 

In Acrostichum infested areas, eradication measures should be 
carried out immediately. It is recommended that this be done manually 
as spraying with herbicides may adversely affect the marine 
environment . 

10.5.4 Nursery operations 

In most cases where flhizpphora sp. are planted, the propagules 
are transplanted to the field immediately and nursery operations are 
not necessary. However, in heavily crab infested areas or areas prone 
to deep flooding it may be advantageous to raise the seedlings in a 
nursery prior to planting in the field. 

For other species such as Sonneratia spp, Avicennia spp and 
Kxcoecaria agallocha, which all have relatively small seeds, raising of 
seedlings in a nursery is advisable. For an excellent account of 
mangrove nursery practises developed for a variety of spscien in 
Bangladesh, please refer to Sidiqi et al. (1993). 

10.5.5 Planting 

In Matang, a reforestation plan is prepared before planting, 
listing the extent and areas to be planted, complete with an estimate 
of supporting resources needed. 


Figure 10 J: Uteophora mcemosa propaguks ready to be planted, Sierra Leone 

Photo by M.L.Wilkie 

Photo by M.L.Wilkie 


The planting spacings are 1.5 x 1.5 m within the swamp for 
Rhizophora apiculata and 1.8 x 1.8 m beside the waterways for 
Rhizophora wucronata. Seedlings are. planted by pushing the radicals 
gently into the soft mud up to about 5-7 cm deep. 

Aerial sowing, on an experimental basis, was tried in the Indian 
Sundarbans mangroves with promising results as elaborated in Box 10*8 


Reforesting mud-ftafc wffli^ iM with a 

helicopter wt* conducted for tbe first time in the toltoa Suodtrtans daring Agu* 
1989, Treated tidai tites comprise regularly f!0M^ 
witb Dbwi gbas grass (Porterosia coarctatt) aad low buafaee 
iUdfrtius). Bacn (Avwr/ww <#d*rfw, A nfer ) and Keora (forwrate 
natural pioneering specie*, were mod, 

Tbe seeding period if from mid Augaat to early September after the ftifl 
moon high tides. With a HILiRVI-EKB helicopter carry^ 30 ba 

were seeded per day at t seeding rate of 6 ha/tour, A total area of 450 ha was 
treated in this way. 

Survival urveys conducted in February ttw fdtowiag year indicated 
establishment rales of 150 - 3*880 seedlings/ha, The seedlings were not umfornriy 
digtributed but with better seed hopper design and using a Beaver type of aircraft 
better results were obtained. 

Source: Lahiri, A. K. 1991 
Box 10.8: Aerial seeding of mangroves 

In Sierra Leone, school children from the local primary schools 
assisted in planting mangroves in connection with the National Tree 
Planting Day. Please refer to Figur*f 10.5 and 10.6 on the following 

10.5.6 Afforestation of newlv formed mudflats 

Extensive afforestation has been implemented, in the Bay of Bengal 
for many years to accelerate the reclamation of newly formed mudflats 
along the coast. Sozweratia apetala and Aviceania officinalis are the 
favoured species (Das and Sidiqqi, 1985). Other species, including 
non-mangroves such as Acacia nilotica, Eucalyptus caznaldulensis and 
Casuarina equi set i folia, which can tolerate saline soils, can also be 
planted on the more consolidated sites. Natural soil subsidence in the 
newly accreted mud- flats can cause some of the plantations to disappear 
overnight. Overall, however, such reclamation efforts have accelerated 
silt entrapment, stabilized new accretions and protective embankments, 
while creating more land that is needed to accommodate an expanding 
rural population and for other agricultural uses. 


10.5.7 Reforestation of degraded areas 

In 1988 a trial planting with 12. harrisonii was initiated at Boca 
Chica, Sierpe-Terraba reserve in Costa Rica, where there were more than 
5 000 ha of Acrostichum infested swampland. A significant proportion 
of the river -banks were denuded. The fern, about 2 m high, was 
manually cleared in rows perpendicular to the waterways. The planting 
spacing used was 1.5 m. x 1.5 m. Low survival rate (about 50%) was 
achieved because the site was rather dry and R. harrisonii might not be 
the suitable species. The seedlings were also attacked by borers. 

Experimental plantings with 12 species were undertaken since 1990 
in the Laputta and Bogalay townships of the Ayeyarwady delta in 
Myanmar. In the elevated and drier sites, Excoecaria agallocha was the 
most promising species. It coppices very well, starts slowly but picks 
up after having consolidated its rooting system. The primary roots 
grow deep into the mud and are therefore, less easily damaged by 
superficial soil cracking experienced during the hot and dry months. 
Some Bruguieras also fair well on similar sites but require better soil 
moisture conditions. The most difficult to re-establish was Heritiera 
femes because it required a good supply of superficial freshwater. The 
network of Heritiera pneumatophores cannot develop properly when the 
surface soil hardens and cracks during the dry months. Sonneratia. 
apetala, . caseolaris and Avicennia officinalis, grow well in the 
wetter intertidal zone that is always moist and regularly inundated. 
Similarly, Rhizophoras grow well on the wetter but more consolidated 
mud- flats that are regularly flooded. Ceriops decandra and C. tagral 
regenerate well under Phoenix paludosa palm under varying degrees of 
shade in the drier soils. Overall, the planting trials indicated that 
the rehabilitation of ex- agricultural lands and denuded elevated sites 
was not easy. Appropriate species must be selected and planted at the 
right time and place, and care must be taken to ensure that sea water 
can reach the planted areas. 

Site, preparation may be needed to reforestate abandoned paddy 
lands located on marginal and drier sites, which can sometimes be 
strongly acidic. The first step is to restore and improve the natural 
soil condition by allowing the Spring tides and rains to saturate the 
soil. This can most easily be accomplished by breaching the man-made 
dikes/bunds that impede water movement. Shallow irrigation canals may 
also be constructed to guide sea water into the planting sites. 
Regular flushing removes toxic soil chemicals and recharges the soil 
with nutrients. Planting of such areas should be undertaken with 
nursery raised seedlings just before and during the monsoon rains. 

Mangrove soils are rich in dissolved salts and pyrite sulphur 
(FeS 2 ) . When the soil is drained and allowed to dry, oxidation occurs. 
Oxidation of pyrites, hastened by bacterial action (Thiobacillue 
ferroxidans) , produces sulphuric acid that reduces soil pH. When the 
soil becomes strongly acidic, sulphate ions react with clay particles 
to release toxic amounts of aluminium ions that inhibit plant rooting 
and even cause fatality - one of the main causes for paddy land located 
in former mangrove areas to be abandoned. 

10.5.8 Refilling 

A year after planting, a survey is carried out to determine seedling 
survival. Areas with less than 70% survival should be refilled during the 
following planting season. 


In most mangrove areas there are practically no weeds. However, on 
drier sites and in more marginal areas there is one important exception: The 
AcroBtichum fern, which in Matang in Malaysia seems to affect even the better 
sites. This fern is difficult to eradicate and in heavily infected areas it 
may be best to do only spot cleaning around newly planted seedlings - 
preferably seedlings raised in a nursery. Manual cutting is advocated though 
the herbicide Velpar is used in some countries. Its affect upon the marine 
environment has yet to be established. 


Crabs may be a major problem in establishing mangroves, as they attack 
the succulent propagules. Various methods have been tried to protect the 
propagules from these attacks such as painting the hypocotyl with yellow 
paint/ placing it inside a bamboo cylinder and planting seedlings instead of 
propagules. However, the most successful (and cheapest) method seems to be 
to let the propagules wither a bit by keeping them in storage for a couple of 
weeks before planting, as this makes them less attractive to the crabs. 

Rhizophora seedlings are sometimes destroyed by stem borers. The 
frequency and severity of such attacks should be evaluated and may be related 
to site factors. In Cuba a large proportion of Rhizophora mangle propagules 
are also attacked by borers. 

In Maswari, Sierra Leone, pristine stands of 35-40 m tall R. racemosa 
were completely defoliated in 1989 by a type of leaf caterpillar, but 
fortunately the trees managed to produce new leaves and survived the attack. 

Herbivore attacks on mangroves in Thailand (mainly by species belonging 
to the Colepptera, Lepidoptera and Diptera orders) are described by Murphy and 
Meepol (1990) and Rau and Murphy (1990) . 

Termites (NaButitermes termitaria) are often found in the cankers that 
develop on the stems, branches and roots above the high- water marks. Infested 
stems become hollow and prone to wind throw. The incidence of termites and 
the fact that isolated trees suffer from bark-scorching, make it impractical 
to retain large stems as seed-bearers, with the aim of producing trees with 
bark of more than 75 mm in thickness. 

In the Sundarbans in Bangladesh, a massive scale of top dying of one of 
the roa^or species (Meri tiara fomes or Sundri) has been reported (See for 
instance BARC, 1990) . Whereas gall cankers are often found on affected trees, 
these infections seems to be of a secondary nature. Increased soil salinity 
has also bean dismissed as the main cause and the general view is that a 
combination of abiotic and biotic factors may be involved. 




^$*j$nk-$'<>* r,:^ 


.^A'l^W^,-' V'..:^.,*. ^- 1 i^^,^'i:- 
.^^^^-/-^M-'? 1 ^'^ 1 '-' 1 ^ 

KV,i5 h ^. i ( i.'',^. *; V'.',' 1 " :' '^"^*,\--"tf ' " 
OT.;'^^: 1 !^.^ V^v^^'jM 
'^^'^V; 1 -' 11 ^-.'^' '^^MJH 






Depending on the growth rate, the length of rotation and the possibility 
for marketing smaller dimension products, thinnings of the stands may be 
conducted to increase the diameter growth of the remaining stock. 

In Ma tang, Malaysia, two intermediate thinnings are carried out when the 
stands are 15 and 20 years old respectively. The method used is called stick 

In the first thinning a 1.2 m long stick is used and a good, straight 
tree is selected. All other trees within a radius of 1.2 m are felled, a new 
tree selected and the procedure repeated. The first thinning thus results in 
a relative tree distance of 1.2 m with regard to the remaining stock, 
equivalent to an average 6 944 trees/ha. The second thinning results in a 
standing stock after thinning of 3 086 trees/ha. The wood obtained from the 
thinnings is sold as poles or firewood. 


There is an optimum size or age to which trees should be grown. The 
period in years required to grow a stand to a desired condition of either 
economic or natural maturity is known as a rotation. It is dependent on a 
number of factors, such as the objects of management, species or combination 
of species put under management, their growth rates,* etc. Environmentally, 
the effects of rotation age on litter- fall should also be considered. 

Rotations may be classified into four broad types as follows: 

(i) The physical rotation; This rotation coincides with the natural 
life span of a species on a given site, which is an important 
consideration for amenity forests, gardens, parks or protection 
forests. In parks that are frequented by the public or in 
managing recreational forests for tourism, moribund or senile 
stems may have to be removed to reduce hazards due to falling 
branches and trees. 

(ii) The silvicultural rotation; This is the rotation where, for a 
given species and site, its regeneration and growth potential 
remains satisfactory. Useful for amenity forests, where a wide 
range of tree sizes and the presence of large, mature individuals 
will enhance the forest landscape value. 

(iii) The ttKrhniffft 1 rotation; The rotation under which a species 
yields the most material of a specified size or specification for 
a special used. To produce fuelwood the rotation period may be 
6-12 years but for charcoal billets varies from 12 - 30 years. 

(iv) The, rotation of the greatest volume production. This is the 
rotation that yields the greatest annual quantity of material. 
This type of rotation is commonly used to maximize production 
when growth increments data are available. The rotation of the 
greatest volume production is the point where the current annual 
increment (CAI) equals the mean annual increment (MAI) . 


y i i ; ' Wtt 1 , ' / . "I" 'V' ' , ' kr 

, ' ' I ,1 ' ' , ' > ' ' . ''f ft "'j"'' '','" ''.''' ; '' 

- ';'.. ! ^> -v.', -.c,.'. v,' ''>, ^'< '.v^'};^" 15 ^.";* 1 '' 1 * ,1 

Figure 10.9: Rhizophora apiculata stand after first thinning, Matang, Malaysia 

Photo by M.L.Wilkie 

figure 10.10: Mature JBMavAow apiculata stand ready for niud fdUng, Matanf , MaUysU 

Photo by ML.Wilkie 


To the above should be mentioned the 'financial rotation' which aims to 
optimize monetary returns on capital under the forest rent principle. Forest 
rent is similar to the culmination of MAI, except that the increment is 
measured in net money received instead of in volume units. 

10.9.1 Selection of Dotation type 

Rotations may be classified into 3 broad groups to meet diverse 
purposes as follows: 

(a) To control the supply of certain services, e.g. for amenity 
forests (Use silvicultural and physical rotations) ; 

(b) To control the production of selected forest products (Use the 
technical and maximum volume rotations) ; 

(c) To control financial returns (Use the rotation of highest 
financial return) . 

Technical rotations are appropriate when a sustained supply of 
timber is needed to meet priority industrial needs or social demands 
for wood. That of a financial rotation works best when the fund for 
silvicultural treatment is a constraint. 

The rotation length depends on the 
following lectors: 

o stand volume growth which 
varies with: 


For Matang mangroves, Watson 
(1928) estimated that the MAI 
culminates at 10.6 m 3 / ha/year at 
39-40 years. If the primary aim is to 
grow trees with bark over 75 mm thick, 
which will normally be found on trees 
of over 30 cm DBH, a rotation of more 
than 40 years is required. Such a 
long rotation may be economically 
difficult to justify, and even 
difficult to apply as in the case of 
the Terraba-Sierpe mangroves in Costa 
Rica due to the high incidence of 
termite and fungal attacks. For these 
reasons, (and due to the waste of 
wood) it is not practical to manage 
forests solely for bark production. 
Bark should thus be produced as a by-product of wood extraction. 

Different lengths of rotation may obviously be selected for 
different species within the same forest, but also for the same species 
according to site conditions, end use, purpose of stand etc, i.e. one 
rotation for each working circle. 

species involved; 

site factor; 

thinning intensity and 


silvics of species such as age 
of seeding, timber quality, etc; 
soil erodibility or deterioration 
after frequent exposures; 
technical factors regarding 
equipment for felling/extraction 

10.10.1 Genetic biodiversity 

A sustained yield system based on a few desirable species or even 
monocultural plantations will reduce the genetic diversity of the 
ecosystem. To avoid this, areas should be set aside for maintaining 
biodiversity. See Figure 10.11 for an example of a pristine mangrove 
area set aside for preservation in Malaysia. 


10.10.2 Erosion control 

The mangrove is a dynamic ecosystem in which tidal and hydrologic 
influences determine the pattern of sedimentation and erosion. The 
body of mud and silt stabilised and held by mangrove vegetation is the 
best form of natural protection. Accretion and erosion of river-banks 
are natural occurrences, and it is nature's way of building up the 
swamp with silt and enriching it with transported nutrients. This 
should be distinguished from man- induced erosion which can be most 
destructive and even irreversible. Destruction of riverine and fringe 
vegetation should be avoided through proper management and control. 

10.10.3 Avifauna 

Migrant shore birds breeding in Siberia, China and Japan use SB 
Asian coastal mangroves and mudflats as resting and refuelling sites on 
their annual migration to Australia. 

The maintenance of Nature reserves where logging is prohibited is 
necessary for avifauna. Herons are commonly encountered at many sites 
of importance to migratory waders. Some build their nests among the 
AcroBtichum ferns. Other species like the Milky Stork Myceteria 
cinerea and the Lesser Adjutant Leptoptilos javanicus seem to prefer 
tall, old trees for breeding (B. gymnorhiza/R. apiculata) . 

Their numbers are reduced due to (a) habitat change/loss (old 
trees essential for nesting are lacking) , (b) hunting (eggs and 
nestlings were collected for consumption) , and (c) human disturbance 
(some sites are regularly disturbed by crab fishermen) . Bird counts in 
mangrove creeks in Matang revealed the importance of the creeks for a 
variety of bird species including the Little Green Heron (But or ides 
striatus) , Common Sandpiper (Actitis hypo leu cos, five species of 
kingfishers and the Masked Finfoot. Some species have feeding 
techniques adapted to foraging during high tide. 

10.10.4 Other wildlife 

Other wildlife such as for instance the Royal Bengal Tiger and 
the spotted deer found in the Sundarbans also require protected areas 
and/or no disturbance in selected areas during their breeding season. 

10.10.5 Fisheries 

As fish and other marine animals use the mangrove areas for 
spawning, feeding and shelter grounds, a belt of mangroves should be 
kept intact along all waterways except for landing and loading sites 
necessary for the felling operations. 

10.10.6 Recreation and education 

Sufficient sites should also be established and maintained to 
enhance the recreational and educational value of mangrove forests. 
Figure 10.12 shows an example of the construction a walkway on stilts 
in a mangrove area designated for recreation. 

All of the above areas may have particular silvicultural needs. 


Figure 10.11: The 'Virgin Jungle Reserve', Matang, Malaysia 

Photo by M.L.Wilkie 

Figure 19.12: Walkway on stilts, CHacap, Indonesia 

Photo by M.L.Wilkie 



11.1.1 Rates of growth 

Growth rates vary, inter alia, with species, site conditions, 
spatial position in the stand, competition status, vigour, and age. 
Notwithstanding, its inherent variability, mean increment per diameter 
size class (where age is not known) is normally used as a measure of 
growth and in stand projection. Growth data may not be locally 
available, but useful indications may be provided by using data 
available elsewhere. (Refer to Table 11.1, presenting data from the 
Matang Mangrove area in Malaysia.) 

Table 11.1: Diameter growth rates of R. apiculata trees by diameter size classes (192041) 

0.26 0.28 0.29 0.25 0.24 

Source: Putz and Chan, 1986 [ * Diameter overbark at breast height ] 

Diameter (cm)* 10 - 20 20 - 30 30 -40 40-50 50-60 

Period measured 1920-81 1920-81 1920-81 1946-81 1975-81 

mean annual 
increment (cm/yr) 

The spatial position of individuals in a stand, expressed in 
terms of crown domination, is useful in determining tree vigour. As 
expected, generally suppressed trees have lower increment rates than 
the dominants and codominant s . (Refer to Table 11.2, Putz and 
Chan, 1986) 

Table 11.2: Diameter increment of Jt. apiculata by crown classes in Pulau Kedl 

Crown class 

Diameter increment (cm/year) 

Source: Putz and Chan, 1966 



Diameter increment data collected by Jimenez, J.A. ( for 

the North Pacific mangroves in the pure Shizqphora forests and the 

overlap zone between Rhizophora sp and Avicennia genninanB are 
summarized in Table 11.3. 


Table 113: DBH (overbark) 

for Rhlzophoras in the pure and overlap 
along the North Padfk Coast, Corta Rica. 

clmmn (em) 
A, 1.7 - 2.9 

Rhiiophor* Zone 
Puxv Overlap 


Rhisophora Zon 
Pux* Overlap 

D. 3.0 - .0 

J.A. (unpublished, 1917) 

11.1.2 Yield and production 

Yitld is the amount actually extracted, while production is the 
total accrued wood increment whether removed or not. The working plan 
should provide a forecast on revenue flows derived from an estimate of 
the commercial stem wood. For unevenaged stands and natural forests, 
where accurate data on growth rates are absent, the potential yield may 
be determined by an appraisal of the current standing volume. 

For example, in Costa Rica an inventory of the Playa Garza 
forests gave plot volumes of 34.6 m* - 373.2 m/ha for stemwood over 10 
cm (DBH.ob.) . The average stand volume was 280.5 m'/ha, which was high 
as the forest had previously been exploited. The mean Rhizoptiora stand 
volume was 163 m*/ha. An indication of the spread of volumes/ha per 
diameter class is shown in Tabla 11.4. 

Table 11.4: Volume In m3/ha per diameter dan, Playa Garza, Corta Rica 

1234SC78 Total pr ha 
10<1S 1S<20 20<2S 25<30 30<35 35<40 40<45 >45 

14.2 22. t 3.0 2i.O 22. < 2S.7 10.7 3.0 1C3.0 St.l 
17. C 27.$ 32.7 17.9 1.4 t.5 4.1 0.0 117. S 41.9 

31. SO. 2 it. 7 45. 9 31.0 34.2 IS. 5 3.0 210.S 100.0 


It was concluded that the final crop would produce at least 150 
m'/ha of JUiizophora on a 25 year rotation, and an equal if not higher 
volume of other species* 

In Ma tang, Malaysia, the average density and volume for 30 years 
old R. apiculata stands are 1 343 trees/ha and 153 m/ha respectively. 
In Ranong, Thailand, an average stocking of 812 trees/ha and mean 
volume of 226 m'/ha have been reported. 

The productivity of plantations in the Upper Gulf of Thailand, 
based on an inventory of privately owned plantations in Yeesarn by 
Wechkit (1987) is presented in Table 11.5. This is not a yield table 
because the data are compiled from diverse sites. 

Table 11.5: Growthaixl Yiddof/M^/w^ 

in Yeesarn, Samut Songkram Province, Thailand 











Stand Density 
No of stow 
per hectare 


cm m 

Commercial Annual 
Volume Growth 
m3/ha nB/ha/yr 












Weechakit, 1987 



The mean annual growth is derived by dividing total commercial 
volume with known age and whereas the table seems to indicate that MAI 
rises significantly after year 10, the data are heavily influenced by 
site factors. 

11,1.3 Effective logging area 

The effective logging area refers to the area actually worked 
after having excluded so-called unproductive areas such as rivers and 
canals, areas required for conservation, coastal protection, research, 
recreation or other purposes incompatible with the felling of trees as 
well as the naturally unproductive sites. The volume yield is then 
calculated by multiplying the effective area to be felled by the mean 
volume per ha. 


11.1.4 Estimation of product -mix 

The mix of forest products to be produced during a plan period 
should be estimated in the working plan, in the form of tonnes of 
charcoal billets, cubic metres/steres of firewood, number of posts and 
poles, etc,. An example of the estimation of product-mix from Costa 
Rica is described in Box 11.1 below. 

fte potential 

dominated stands were determined 

estimating die naniber of 1,5 m sectional lengths obtainable from merchantoWe-stem 
wood op to tO cm uadwrbark tor <8flteit*t diameter size-classes, 

The ttOaaoaihip between mercbiwttWe height <Hm) to 10 em top diaineter undertwk and 
was given by the tinmr regression pirn =-6,564 -K^.864<DBH)] with a correlation 

TWs regrcoon wa uaed to predict 
the awctantobte height <i.^ merchantrf>Je bole) for different diameter ri^e ,ctaue& 
Knowii* this, ltepoieiW mix of end-products determined based on the known product 
specificatttw and dfcrir martow values for a given country, as for example shown forCotta 
Rica in TMe 11J md 11,7 below. 

potes can also be used for construction purposes. la Costa Rica, 
the tttaiards used for wooden potes are based on that for Western Red Cedar and 
Ponterosa Pioc instead of local tknber. Locally, concrete petes are also used and these 
we considerably more expensive dun wooden ones* 

Chang (198a) 
BoxlLl: Efittinatkm of product-mix in CoataRica 

Table 11.6: General specifications for selected mangrove-based products in Costa Rica 





1 Pirwood 







- 25 


Fencing pomtm 


- 21 

2 - 2 


Telegraph pol 


- 21 


Tranwaiaaion polaa 


- 31 





4 - 



cm overbark) 

haight () 



11.2 Forest Yield Regulation 

A regulated forest is one in which sustained yield condition operates 
over all parts of the managed property. This is seldom possible in practice. 
It is determined by, (a) the rotation period, and (b) the annual cut. 

11.2.1 Determination of the annual cut 

All silvicultural operations culminate in the removal of forest 
products, such as fuelwood, poles, posts and timber, etc. The final 
felling has the most decisive influence on the forests and is in itself 
a silvicultural operation. The determination of the type, location 
and amount of the cut is crucial, therefor^, for the future shape and 
development of the forests. 

A cutting policy determines the following, viz., how much to cut, 
the kind, quality and dimension of produce to be harvested and where to 
cut and in which sequence? It is guided by following considerations: 

(a) achieve management objectives; 

(b) market situation for different products; 

(c) silvicultural needs and constraints; 

(d) harvesting constraints; 

(e) environmental impact on non-wood values, e.g., fishery, 
mollusca, apiculture, wildlife and ecotourism potential which 
may be disrupted by extensive clear-felling; 

(f) social aspects such as sustained employment, off-season 
employment opportunities, etc. 

There are two possible approaches to determine the cut; viz., (a) 
by means of area control, and (b) through volume control. 

Neither approach, can by itself, be completely satisfactory 
because a volume to be cut is meaningful only if it is location- 
specific, otherwise is difficult to apply and supervise. Thus, area 
and volume approaches are complementary and often combined. 

Area Control 

The principle of area control is that annually a certain area of 
forest is available for final felling. Where the site is very variable 
and/or the area is large, it is usually not possible to stabilize 
volume production based on cutting an equal area of forest annually. 
Fluctuations in yearly volume yields can be mitigated by the device of 
using ecu ior oduc t i ve or 'reduced' areas, which take into account site 
productive capacity. 

Examplei If a uniform forest is managed under a 30 years rotation, then each year 1/30 of the 
total forest area reaches the rotation age and will be harvested and regenerated. This can be 
expressed by the following formula: 


AC - A/R where AC - Annual Cut in ha/year; 

A Total productive forest area in ha. 
R - Rotation in years 

Note: The Annual Cut in this example is the area annually available for final felling. An 
equal size area is available for each of the prescribed intermediate thinnings. 

Volume Control 

In this method, the cut is determined by the volume and 
distribution of the growing stock and/or the increment. The required 
data are derived from forest inventory results. Von Mantel's formula 
which is entirely based on the growing stock may be used as a guide. 
The formula may be expressed as follows: 

AC - 2(Gs/R) where AC Annual cut (m 1 ); 

Gs Growing (standing) stock (m*); 
R m Rotation (years) . 

Example : Given a total forest area of 3 000 ha with a standing stock of 360,000 m 1 and a 
rotation of 30 years, the annual cut in m* will be determined as follows: 

AC 2 x (360 000/30) 
= 24000m a /year 
- 8 m*/ha/year. 

Note: This estimate of the allowable annual cut in m'/ha is inclusive of the volume 
removed through intermediate thinnings. 

The value of Von Mantel's formula lies in its simplicity, the 
small amount of data required for its use, and its usually rather 
conservative results. 

However, it can only serve as a quick and rough approximation, 
and its application is restricted to forests of even- aged stands with 
a balanced distribution of age classes, i.e. a 'normal forest'. In 
practice, this is seldom the case. 

More precise control methods incorporate the annual increments 
and combine the area and the volume control methods, as these two 
methods are complementary. For this combined approach the data 
requirement is more demanding and may be difficult to fulfil. 

As a general rule for plantations with comparatively short 
rotations, the area control method will provide acceptable results, 
whereas in natural forests with uneven- aged stands the volume control 
method will lead to better results. 

Detailed coverage of the more sophisticated control methods is 
beyond the scope of these guideline. Reference should be made to 
standard forest management texts, (e.g., Davis, K. 1966; Osmaston, F.C. 
1968, Clutter, JL. at al, 1983) 


11.2.2 Regulation in eveq-aoed 

Stands in which the dominant and co-dominant trees are about the 
same age are defined as "even-aged 11 . Even-aged stands may be 
established as follows: 

Afforestation or reforestation after clear-felling; 

Coppicing system; 

Seed tree method and natural regeneration (shelterwood system) 

However, by relying completely on natural regeneration it may be 
difficult to achieve even-aged status, unless steps are taken to 
intervene either by spot, enrichment or even block planting when the 
regeneration stocking is judged to be inadequate in numbers or unevenly 







25 -YEAR CUT* 

20-YEAR CUT-^ 





^yyyip ]f 




x^^ r ! ! ROTATION 


v^ 4 


) 5 10 15 20 25 30 TOTAL 

Figure 11.1 Life-pattern of an even-aged stand 

The management of even-aged stands is characterised by the 
application of clearly defined rotation periods and silvicultural 
measures aimed at achieving an even age-class distribution. 
Figure 11.1, adapted from Davis (1966), graphically depicts the life- 
pattern of an even-aged stand over a rotation of 30 years. 

11.2.3 Regulation of uneven- aoed stands 

The Society of American Foresters defines an uneven- acred stand as 
one in which there is considerable difference in age of trees and in 
which three or more age -classes are represented. 

The management of uneven-aged stands is directly related to a 
certain cutting cycle which starts with a certain reserve of standing 


The cutting cvcle is the planned interval between malar felling 
operations in the same stand, and is determined by the stand volume 
increment. Felling begins once the growing stock has attained the 
desired volume or the desired dimensions. A number of marked or 
selected stems are removed the volume of which in aggregate equal the 
volume of the total increment of the stand within the cutting cycle. 

The general objective is to harvest before the increment of the 
stand volume decreases significantly and to maintain a defined reserve 
of growing stock. Figure 11.2 illustrates the effects of the cutting 
cycle on the growing stock of an uneven-aged stand. 




Figure 11.2 Ufe-pattern of an uneven-aged stand 

11.2.4 Control of removals 

Due to the variability between stands, it is impossible to 
extract equal volumes every year, therefore, a 5-10% variation in 
annual yield can be expected. In order to control the yield, the 
planned cut is compared to the actual cut through periodic inventories. 
The purpose is to maintain a sustained yield of forest products. 

It is clear that only in ideal situations will the actual annual 
cut corresponds to the one which was determined at the beginning of the 
inventory period. In principle, however, the total actual cut or 
removal in volumetric terms within the planning period should 
correspond to the planned one, i.e., the total cut divided by the 
number of years of the planning period should result in the planned 
annual cut. This can be controlled by using the following formula: 

- V t ) 



(m /ha/year) 

Where : 

1 mean annual increment during the period in m 1 /ha/year; 

V t + n - stand volume in m'/ha at the end of the planning 

V t stand volume in m j /ha at the beginning of the 
planning period; 

Hn sum of the intermediate removals in m*/ha within the 
planning period; ' 

n planning period in years. 

Sustained yield management would be judged to be prevailing if 
the total actual cut during a plan period does not exceed the 
accumulated annual increment during the same period. 

Example : 

V t + n - 180 mVha; 

V t 120 m'/ha; 

Hn = 30 m'/ha (in the form of pole thinnings); 

n ~ 10 years (working plan period); 

Substituting, therefore: 

(180- 120) + 30 
I s 9mVha/a 


In this example, the mean annual increment of 9 m 1 /ha/year is 
slightly higher than the average annual cut of 8 m /ha/year as shown in 
the earlier example. The growing stock has increased and sustainable 
yield is achieved. 



The felling plan sets out the sequence of logging operations over space 
and time, complete with felling prescriptions to guide the area forest 
manager . 

Preparation of a felling plan begins from the design of inventories 
which should ask basic questions such as: Where, when and how the 
cut ting/ Togging operations tafce place? (See Tabl* 12.1) . A felling plan must 
contain a clear statement showing the average annual cut in for a fixed 
duration, say the next 10 years. Clear prescriptions are given to achieve the 
production targets and forest renewal programmes planned. 

Table 12.2: Example of detailed objectives for a Felling Plan Inventory 

Part of the General 

Detailed Objective 

Data specification 


1. Mapping of stands, compts, and sub-compartments 
(treatment units) 

Aerial photography 
Field survey - maps 

2. Description of stands and compts. with regard to terrain, 
stand and tree factors 

Field work. Integrate 
data from strategic 
inventories (data 

When and how should 
the cuttingUogging 
operations take place? 

3. Operative and relevant classification of stands and 
compts. from management system point of view, priorities 
with the 10 - years period and the cutting/logging system to 
be applied. 

Accessibility studies 

Source: N.E. NUtson. 1971 

12.1.1 Fellino strips 

The width of the felling strip is often set to 50 m which is 
about 1.6-2 times the height of the predominant trees. It can be 

Orientation of felling strips 

Felling strips are normally oriented perpendicular or 45 to the 
out-going tide to facilitate tidal flushing, promote seed dispersal, 
reduce insolation and provide shelter from prevailing strong winds. 
Generally strips should not be inclined towards the direction of river 
flow near the river mouth because of the danger of excessive silt 
deposition during peak flows. Further way from the coasts into the 
more sheltered estuarine areas, the orientation of the strips becomes 

less important. Here the main rnnrern -in fn f ari "H I-AI-A ffirionh 


A harvesting system is the coordinated package of activities and methods 
used to fell trees, snig and haul logs and transport them to market. The main 
constraints in choosing a harvesting system may be the climate, soil, terrain, 
equipment available as well as the need to protect the site and residual 
trees. Ideally, the chosen system should yield the lowest cost per unit 
volume at the forest gate or of timber delivered to the market place without 
impairing the site and growing stock. 

Planning a tidal swamp harvesting system is a complex undertaking, as 
the available options are limited. Until recently, the difficult ground 
condition has always precluded wood extraction by any but primitive manual 
methods. The low unit value of the produce also limits the use of cost 
intensive systems, while the need to reduce surface disturbance implies that 
precautionary measures and related costs must also be included. There is, 
therefore, a gulf between what is silviculturally desirable and what is 
economically and technically feasible. 

Each silvicultural system produces a distinctive form of crop which 
generates particular problems in felling and extraction. A uniform system may 
be more cost effective as larger amount of wood is extracted per unit area. 
At the same time, the products are usually more uniform in size and amenable 
to handling. A system which retains a significant number of seed-bearers will 
not be suitable for high- lead cable extraction for instance. 

In Asia, generally, forest operations are very labour intensive. 
Increasingly, the extraction industry is facing difficulty in hiring skilled 
manpower. In West Africa, wood extraction is usually not properly organized 
as the supporting processing industries are often not in place. In Central 
America, until recently, the emphasis has been on bark collection rather than 
charcoal billets and other products. In Venezuela, a barge equipped with an 
A-frame cable system was used to extract Avicennia sp for mining props, 
railway ties, and telegraph/transmission poles. Such operations are feasible 
only in mangroves with deep waterways. Removals must also be large enough to 
justify the capital outlay required in such mechanised operations. In Cuba, 
where there is a shortage of railway ties for the sugar-cane industry, 
Avicennia germinans logs are extracted through a system of artificial canals. 

At the initial management and silvicultural conversion phase, the forest 
may contain many over-mature residuals, which must nevertheless be felled, 
removed and used, if the chosen silvicultural system is to succeed. This 
first crop may thus pose special problems whereas in the succeeding crop the 
dimension of the forest produce will be more uniform and manageable. 

A capital intensive system should be avoided in countries where, 
generally, a policy of import substitution and conserving foreign exchange 
prevails. Instead, simple but cost-effective systems should be adopted, which 
can be progressively upgraded as local servicing support facilities are in 
place. Along the coastal zone, the general shortage of mechanical support 
services and the high cost of transporting replacement parts to the mangroves 
are additional constraints to be considered. Furthermore, due to the 
corrosive saline environment and tidal nature of the mangroves, mechanical 
equipment requires close maintenance. 


The main methods used for extracting mangrove wood products are: (a) 
wheelbarrow, (b) tramway, (c) canals, (d) high-lead cable, (e) portable cable 
winch and (f) manual. A brief review of these methods is given below. 

12.2.1 Wheelbarrow method 

This method is still used in Matang. It is labour intensive. 
Wooden planks (1" thick x 9" wide) about 5 m in length are laid across 
the felling coupe. Billet or firewood loads of about 300 kg are 
manually pushed to the boat landings, using locally made wheelbarrows 
over an average distance of 150 m. A shoulder strap is often used to 
help lift and balance the wheelbarrow. The planks are replaced about 
every six months. Both the wheel and axle of the wheelbarrow are made 
of wood to resist salt corrosion. This method is suitable for the 
removal of billets (1.6 m long) and is not used for the removal of pole 
thinnings. This is a simple, practical and low cost method. It may 
not be suitable for frequently flooded areas because the planks may be 
washed away by the retreating tides. (See Figures 12*1 and 12.2) 

The billets are loaded manually on to the boats. Two or more 
boats are towed by a diesel powered boat of 25-35 hp. Each boat can 
carry about 3 t. In Thailand as many as 12 boats are towed by a single 
tug boat. 

12.2.2 Tramway 

Unlike the Rhizophora dominated stands, which are situated in the 
lower intertidal zone and are encumbered with tangled stilt roots, 
forests in the elevated inter- terrestrial zone are dominated by species 
which, generally, do not have prominent aerial roots, such as Bruguiera 
gymnorhiza and B. caryophylloides . These trees can grow to large sizes 
and, as they are situated in less frequently inundated areas, they can 
usually be accessed by light trolleys or small-wheeled carts on wooden 
rails or tramways. 

12.2.3 Canals 

Extraction canals are used in many parts of the world, notably 
Malaysia, Vietnam and Cuba. In Cuba, the early Spaniard settlers dug 
canals to extract Yana (Conocarpus erectus) for making charcoal. 
Though there are reservations that its use may be environmentally 
undesirable because local micro-relief and ecology may be significantly 
altered, no quantifiable adverse results have been reported. Canals 
aligned parallel to the felling strip (50 m x 200 m) facilitate the 
rapid and orderly removals, thereby reducing disturbance to advance 
growth. During logging damage to seedlings and saplings is normally 
high. Most of the younger and less lignified Rhizophora regeneration 
will coppice, and can also straighten themselves into an upright 
position if pushed over and held down by debris. Logs, posts and 
billets are loaded onto the boats usually at high- tide. 

Canals are constructed manually, mechanically, or with 
explosives. In Matang canals are dug by hand. Billets are manually 
carried and/or wheelbarrowed to the stacking and loading sites. 
Felling is carried out by a combination of chains aw and handsaw/axe. 


Figure 12.1: Extraction of wood by wheelbarrow, Matanf , Malaysia 

Photo by M.L.Wilkie 

Figure 12.2: Loading of billets unto the boat, Mating, Malaysia 

Photo by M.L.Wilkie 


Small boats are guided up the canal during high tide and loaded 
manually, then taken out of the canal and normally rowed or sailed to 
the market at collection points. Motorized boats are now used. It is 
very labour intensive and well suited to rural settings where labour is 
plentiful but jobs are limited. 

In the Ca Mau peninsula in Southern Vietnam, the main connecting 
canals were made by dredges and some of the Forest Enterprise still 
operate such dredges. Most of the minor extraction canals were 
constructed manually. These may be temporarily adapted for shrimp- 
farming in certain cases without disturbing the environment. 

In Guana 1, Cuba, extraction canals are constructed with 
explosives (Amonita GJB, Amonal, Roca amonita y Nitromiel) . It is very 
rapid, labour extensive and cost effective. Canals are made under the 
supervision of explosive experts provided by the local militia. Unless 
properly planned and implemented, this method can be ecologically 
disruptive, particularly to wildlife during their breeding seasons. 
Similarly unless open-ended canals are made connecting two bodies of 
water, a localised increase in the salinity may be induced during the 
summer months. It should be noted that mechanical methods are no less 
disruptive and are capital intensive. See Figure 12.3 below. 

Some guidelines on canalization based on the Guanal area are 
summarized in Box 12.1 on the following page as an example of what 
precautions should be taken to avoid negative environmental impacts. 

Rfure 12 J: Mn-Mde CUM! In Gwnal, Cuba 
Photo by P.W.Chong 


The existing networt of primary and secondary canali, including tbcir 
navigable wkbh and operating depth should be surveyed and mapped; 

Before contraction, the canal should first be lurveyed. ttihoutt be aligned 
towards the pratorainatiflg winds to promote better seriate wtter flow 
tod oxygeaatiott 

Tfee principal canals should be interlinked at bob ends to waterways; 

middle lection will be exploded first with both ends of the canal intact 
and closed to reduce tile disturbance and advene environmental impacts; 

When the water in the exploded section has cleared, the ends can be 
breached preferably during low tide, BO Oat adds may be flushed out 

Canal depth tbouM not exceed 1,5 m deep to minimize the quantity of 
overburden. Excetsive toil removal increases the risks of acid-tulphation, 
impedes water movement, covers up too much of fertile surface soils and damages 
advance growth, 

Explosives should not be used close to wildlife nesting or resting sites, 
particularly during their resting/hatching periods (February-April and April- 
June are the respective nesting seasons for the Cuban and American crocodiles. 

An environmental impact assessment in selected areas should be conducted 
to particular, the impact of canalization on local site factors (sofls/saKnity/pH), 
water movement, flora and fauna should be determined. 

Source: Chong (1989b) 
Box 2.1: Guidelines for canalization with explosives in Guanal, Cuba 

12.2.4 Hiah-lead cable system 

In the San Juan-Guarapiche area, in Venezuela, harvesting of 
Avicennia stems and poles is organized into clear- felling strips 
perpendicular to the river bank. The strips average 50 m in width and 
up to about 300 m in length. A high- lead cable system mounted on to a 
barge towed by a tug-boat is used to haul the timber to the 
stacking/loading sites along the river bank. Within each clearcut 
strip two cable settings are typically used so that the maximum lateral 
reach of the cable system is limited to 25 m or less. This reduces the 
damage to the soil. 

The wood is used for telegraphic/telephone posts, mining pit- 
props and general utility timber. Logs and poles are transferred onto 
barges and transported to the jetty. 

Small boats are also used to transfer the timber to the barges, 
especially when the stacking sites are situated along shallow creeks. 
The barge has a capacity of 400-450 pieces of poles or 18,000-24 000 kg 
of billets. In Venezuela, long-distance transport to the processing 
mills is carried out by floating platforms hauled by tug boats. 


The high- lead cable system is a capital intensive system and 
therefore can be used economically only in areas with high volume of 
commercial timber and/or where the wood has to be transported over long 
distances. Additionally, it can only be used for clearcutting as the 
action of the high- lead cables will destroy any remaining trees within 
the area. Furthermore, the rivers and creeks have to be deep enough to 
permit the use of shallow draught barges. 

12.2.5 Pgftfflble cable winch 

In Costa Rica, a portable winch powered by a small chainsaw motor 
Has proved to be a useful alternative in timber harvesting. Stems, 
poles, firewood and charcoal billets are hauled with minimum 
disturbance to advance growth. It was first tested in the Sierpe- 
Terraba mangrove reserve along the Pacific south coast. A strip clear- 
felling method was used. The ultimate goal was to transform the 
irregular Rhizophora dominated forests into a series of even- aged equi- 
productive stands. During the conversion phase, the existing forest, 
which has been selectively logged, has a high proportion of over -mature 
residuals, which must nevertheless be felled, removed and utilized, if 
the chosen even-aged silvicultural system is to succeed. 

Manual removal of large sized individuals was difficult as the 
soil was soft. The diameter ranges from 10 cm to over 45 cm. Because 
of its portability, a light winch could easily be moved from place to 
place. It was found to be very practical, low costs and easy to apply. 
A load of billets attached to a skidding sleigh of 1/2 t can be hauled 
without difficulty over a distance of 200-250 m. Training was simple 
because most of the locals knew how to operate a chainsaw. 

With this method it was found possible to remove the large sized 
trees which were often left by the bark collectors, who only exploited 
the bark and not the wood. 

Under Cuban conditions, working in the Avicennia germinans 
dominated forests, it was estimated that the average snigging distance 
using an experimental portable winch was about 60 to 80 m, and 20 - 30 
m with manual methods. With a larger winch and longer cables, it may 
be possible to reduce the intervals between extraction canals very 
significantly. On this basis a cost comparison between tnanual and 
winch methods was made and shown in Table 12.2 below: 

TabteUJ: Comparative mining costs urinf manual and winch methods 

Swigging *thod Manual 

ioure*: Chong, 19M/ lugonio Pouiiin Nolinet, IMS 

Dtnaity of canal* (/ha) 
OMt of canal* (po/ha) 

12.2.6 MflflUB 1 ext^ rac t i on 

in Sierra Leone, the trees are cut using a local type of axe. The 
smaller firewood billets are normally bundled and carried manually to 
the dug-outs. 


Extensive tests carried out in Mali show that a labourer is able 
to transport one stere of firewood over a distance of 100 m in about 
1.2 hours. (Atlanta, 1986). With regard to the bigger billets, most 
workers will not carry a load of over 150 Ibs, and then only over short 
distances . 

Directional felling is possible with proper axes, handsaws and 
wedges. Removal of billets may be facilitated by the deliberate 
felling of convenient trees in such a manner that the topped trunks lie 
end to end forming a rough track across the swamp. 


In the following the main causes of logging damage in relation to the 
choice of harvesting system are described together with mitigating measures 
to be taken based on Hamilton and Snedaker (1984) . The environmental impacts 
of mangrove management per se is the subject of Chapter 14 to which the 
reader is referred for further details. 

The harvesting and extraction of wood products always causes some 
adverse effect on the forest soil unless the logs are extracted by high- lead 
cable without touching the ground. 

Dragging of trees and logs result in removal of the top soil, damage to 
natural regeneration and remaining seed-bearers and compaction of the soil. 
The latter is also causes by using tramways and wooden rails. The skid rails 
may later be prone to deep flooding and thus have a negative effect on natural 

Construction and widening of canals alter tidal influence and draining 
patterns . 

As is evident from the above, all of the harvesting systems described 
have advantages and disadvantages and none of them can be universally applied 
due to the differences in working conditions between countries. In order to 
minimize logging damage, the following should be kept in mind when choosing 
and implementing a particular harvesting system: 

Select the most appropriate harvesting system keeping in mind the 
environmental impacts of each system as well as their technical and 
economical feasibility under the local conditions; 

Design extraction methods to minimize damage to seed trees and advance 
growth. Such measures include directional felling, proper lay-out of 
loading site and skid rails, the use of two barge settings in each 
felling strip when the high-lead cable system is used etc.; 

Adjust strip width in clear fell ing to ensure supplementary regeneration 
from adjacent uncut areas; 

Retain buffer strips along rivers and waterways to stabilize banks and 
as a habitat for birds and marine life; 

. Attempt to reduce slash size and volume. A large amount of slash 
represents not only an under-utilized wood resource, but a damaging 
agent to natural regeneration until it has decomposed. 



The operational plan can either be a working plan covering all the 
aspects of forest management implementation within a forest district or a 
range limited in time to 2-3 years or it can consist of specialised plans for 
each of the major operations such as a silvicultural plan and a felling plan 
often prepared for the full plan period (10 years) with more detailed plans 
developed for each year. 

Such plans specify the location and the operations to take place in 
details and include information on organisation of work, man hours and 
equipment needed, costs and yearly targets. The operational plans are used 
to guide the day to day implementation of the overall forest district 
management plan. 


13.2.1 Records to be kept 

In order to ascertain that the objectives of the management plan 
are met a set of records are kept where the actual achievements are 
registered together with the annual targets. 

For silvicultural operations such records contain information on 
size of area planted, number of seedlings /propagules per ha, manhours, 
costs, survival rate 1-2 years after planting etc. For felling 
operations the records show the size of the area to be felled, expected 
and actual yield, manhours, costs and revenues, natural regeneration 
status etc. 

13.2.2 Supervision and control 

Supervision of activities is carried out by rangers, forest 
guards and other forestry staff. However, some aspects may be cover by 
contractors . 

Felling activities 

In the clear- felling in alternate strip method with or without 
the retention of standards, the contractor should carry out the 
following before logging begins: 

after the forest ranger has shown the logging area (coupe) and 
its boundaries to the contractor, he must demarcate the 
boundaries satisfactorily before logging can begin; 

all seed-bearers are identified and marked on the instruction of 
the area forester; 

he should furnish a list of his subcontractors or workers engaged 
to work in the licence/permit area; 

the felling coupe is subdivided into two more felling blocks and 
it is the contractor's responsibility to be fully aware of the 
layout of the blocks. 


The range forester should prepare a felling plan map at 1:5 000 
or 1:10 000 scales, noting the following: 

storage and loading points; 

gaps with good regeneration where directional felling is needed; 

excluded areas such as research plots, etc. 

progress of felling blocks. 

Control bfttwttn falling strips 

Felling should be conducted in an orderly and sequential manner. 
Normally, the contractor is not permitted to extract timber in the 
second and subsequent felling strips, unless at least 80% of the 
preceding felling strip has been properly worked. No more than two 
felling strips may be worked simultaneously. 

Control within falling strips 

The effective area of each felling strip is subdivided lengthwise 
into felling blocks at about 100 m intervals. Felling is permitted in 
the first two blocks, but may not be permitted in the third and 
subsequent blocks unless one of the preceding block has been "closed". 
A "closing report" is prepared by the range officer based on field 

Marked standards should not be damaged or felled and logging 
slash to be properly disposed. If such standards are accidentally 
damaged, an alternative tree should be substituted. Over-mature or 
moribund stems are felled and removed. 

Control of rtaovals 

To regulate removals, the movement of the produce has to be 
supervised and controlled. 

In Matang only boats registered with the Forest Department are 
allowed within the Forest Reserve and permitted to operate and carry 
forest products. The same is true for the Sundarbans in Bangladesh, 
where the Forestry Department also is in charge of issuing licenses for 
fishing and honey collection within the area and thus has control posts 
situated at all the waterway entry points to the Reserve. 

Silvicultural activities 

Most silvicultural measures in mangroves, such as planting, 
weeding and pest control, are carried out by forest staff under the 
supervision of the forest ranger. 

13.2.3 Costs and revenues 

An essential objective during the plan period is to quantify 
costs of forest establishment, harvesting and administration. Bach 
forest range should be treated as a cost centre. 

Cost records are similar to those kept for inland forests and 
vary according to countries and will thus not be described in detail. 

Revenue collection 

There are three methods for revenue/royalty collection, viz. , (a) 
collection at stump, (b) at wood processing sites, and (c) in transit 
by means of checking stations. 

The first method is recommended because it requires few staff and 
it encourages the contractors to be more efficient in wood harvesting 
and extraction. However, this method requires quite intensive 
inventories of the stands prior to felling. 

.In Matang, Malaysia, the premium for wood for charcoal is paid 
prior to the commencement of the felling operations based on a 
inventory of the area to be felled. The royalty on the other hand, is 
collected once the charcoal has been produced at designated processing 
sites. Royalties for firewood, post and poles are collected at source 
in the form of prepaid licences or through the issuance of removal 
passes validated at approved checking points. 


Ideally the forest management plan should be evaluated at least once 
during the plan period and revisions incorporated where necessary. However, 
due to the amount of work involved in such an exercise, plan evaluation is 
often only undertaken in connection with the preparation of the next 
management plan. 

Such an evaluation should inter alia encompass an assessment of the 

standing stock and growth rate compared with the estimated production 
and the actual yield; 

the environmental impacts of the current harvesting system and an 
examination of mitigation measures which could be undertaken; 

plan objectives; 

the need for changes in current silvicultural operations and 

further research needed in order to refine the present management 
prescriptions . 



There are several environmental concerns regarding the use and 
management of the mangrove ecosystems. The level of concerns, however, vary 
from country to country and from region to region in the terms of severity and 
areas of impacts. The concerns are most severe at the local level where 
remedial measures are much needed. 

The most common concerns worldwide are: 
(i) Deforestation of mangrove forests; 

(ii) Depletion of fisheries and other mangrove dependent living 
resources ; 

(iii) Loss of protective function of mangroves, in particular in 
coastal areas subject to severe storms and wave actions; 

(iv) Serious degradation or destruction of critical habitats; 
(v) Loss of bio-diversity/genetic resources. 

At the local level, the concerns are directed at more specific losses 
of production and production potentials and services (see Part II) . 


The questions to be asked are whether mangrove forests can in fact be 
managed on a sustained basis and if so, does the management have any 
significant detrimental effect on the environment. To the first question we 
have to rely on the experience from two areas that has been under management 
for a long time, the Matang Mangrove Forest Reserve in Perak, Malaysia, and 
the Sundarban mangrove forests in the Bay of Bengal. 

The Matanq Mangrove Forest Reserve has been under intensive management 
since the beginning of the century. The forest is regulated in the 
appropriate even- aged classes and harvested stands are immediately replanted 
and, to a limited extent regenerated naturally; afforestation of marginal 
areas incorporated into the reserve is generated exclusively by means of 
plantation. The preferred species is Rhizophora apiculata although other 
species are also utilized. 

With regard to the 5 concerns listed above, it is relevant to ask if 
this form for intensive management of mangrove forests, which has been widely 
regarded as an ideal model and in fact copied or adapted to local conditions 
in Thailand and elsewhere, has been sustainable; and whether it has any 
negative effect on 

fisheries and other mangrove dependent living resources; 

critical habitats; 



Monitoring of the permanent sample plots installed in the reserve has 
not shown any decline in yields over the last 3 rotations (Haron 1981) . The 
incorporation of marginal areas by canalization and plantation has increased 
the size of the management area over the years. It must, therefore, be 
concluded that sustained yield forest management is feasible at least for 3 
rotation periods of 30 years. 

With complete crown cover and immediate regeneration of harvested areas, 
the production of litterfall from the managed forest has been found to be 
superior to that of natural untouched stands (Ong et al., 1982). Litterfall 
is an important source of energy for the food web that affects fisheries and 
other living mangrove dependent organisms. There has not been any significant 
decline of commercial catches of fish, mollusc and shellfish in the coastal 
waters adjacent to the Matang Reserve reported during the management period. 

It can not be ruled out that critical habitats have been modified or 
radically changed in Matang, although no reports in this respect are 
available. Marginal areas for wood production have been provided with canals 
in order to lower the salinity levels to those appropriate for Rhizophora 
apiculata and Rhieophora mucronata and other desirable tree species, and in 
this process it is possible that critical habitats have been disturbed. 

It would seem most likely that the original bio-diversity of the 
original mangrove has been affected in the regulated parts of Matang since 
only a few preferred species have been used to regenerate harvested areas. 
Consequently, with this type of management, areas should be set aside in the 
initial planning stage (or at the Integrated Coastal Area Management planning 
exercise) as reserves managed for there genetic resources rather than for 
commercial wood production. 

In the Sundarban mangroves, management has been practiced for a much 
longer time and developed along different lines than those of the Matang 
reserve. The fundamental differences are, that the total area under 
management is almost 10 times as large; many more species are utilized - sawn 
wood being a major project; and the shelterwood system is applied to promote 
natural regeneration. 

Although the regeneration system should maintain bio-diversity, the 
inherent complexity of the system also makes it difficult to assess, if this 
has in fact happened. However, one would assume that it has: But the most 
important advantage of the forest management of these mangroves is, that they 
have been retained at all. As has been the case elsewhere in the tropics, the 
Sundarban mangroves have been under heavy pressure to be converted to food 
production, i.e., fishponds, shrimpf arming and agriculture. The fact that 
there are such an extension of mangroves still in existence today, in two 
countries with some of the highest population pressures in the world, is 
largely due to the fact that a significant part of them were reserved at an 
early stage for forest management. 

Nowhere in the world are mangroves more important for protection of 
human lives and activities than in the Bay of Bengal, which is periodically 
affected by dangerous typhoons. Consequently, it can be safely stated that 
the forest management of these mangroves has had a positive impact on the 
environment . 


Natural watcrborne feed is also still available for fish and shrimp 
farmers in the Sundarban contrary to some parts of the world, where heavy 
deforestation of mangroves, not subject to management, has made it necessary 
for the shrimp farmers to resort to artificial feed in recent years (Ecuador) . 

The use of inappropriate management techniques can, however, have a 
detrimental effect on mangroves resulting in deforestation, losses of critical 
habitats and bio-diversity. The harvesting of mangroves for chip production 
using large and costly machines, has made it economically necessary to 
clearcut large areas (Indonesia, Sabah and Sarawak) . The regeneration of 
these areas have either failed or proven to be very difficult and this type 
of operations have, at least in part, been stopped. For more details on 
logging damage in relation to various harvesting systems please refer to 
Chapter 12. 

Canalization of mangroves to facilitate wood .extraction and allow sea 
water to enter into non productive salt flats are beneficial if properly done. 
However, cases have been observed where canals constructed with explosives 
were banked in a way that did not permit water to flow freely in the affected 
area. This in turn led to localized high salt concentration patches with 
stunted tree growth. 

In conclusion, it would seem that forest management of mangroves, if 
properly carried out, has a largely beneficial effect on the environment. 
However, integrated management planning should ensure that all goods, services 
and values are catered for. In order to ensure that is in fact done, it is 
recommended that the management planning includes an Environmental Impact 
Assessment, and that actual management is monitored periodically by 
Environmental Auditing. 



These conclusions, aimed at achieving sustainable mangrove management, 
are drawn from regional and selected country experiences. Accordingly, they 
are general in nature. Similarly, the guidelines on "recommended practices" 
highlighted in italics should also be adapted to suit local requirements. 


15.1.1 Land use policy 

To achieve the desired management goals, an important 
precondition is to redefine and/or prepare a land use policy so that 
mangroves are recognized as a legitimate form of priority land use 
rather than a residual use. 

A national land use policy to ensure the sustainable use of 
mangrove resources, and providing for the establishment of a permanent 
resource base, should be redefined and/or formulated and implemented. 

(a) Proposals for converting mangroves to non-wood uses should be 
supported by favourable environmental impact assessment reports. 

(b) Solar salt should be produced only in the semi-arid or seasonally 
dry zones, and be located on natural salt-flats with restricted 
or sparse natural vegetation cover. 

(c) Potential acid sulphate and acid sulphate mangrove soils should 
not be cultivated under rice or other crops, without an adequate 
source of rain- fed or irrigated freshwater. 

(d) Tin-mining activities should be avoided along the coast and the 
mining discharge should be confined to sedimentation ponds and 
not discharged directly into the river system; 

(e) Aerial spraying of pesticides and chemical fertilizers adjacent 
to the mangroves should be avoided. 

15.1.2 Multiple-use concept 

Single-use management should be avoided as this forecloses the 
many direct and indirect benefits and services that the natural 
ecosystem can offer on a continuing basis. 

A policy statement on the multiple-use management of mangrove 
resources, particularly for forestry, fishery and wildlife conservation 
should be formulated and politically supported at the highest level of 

To ensure an integrated approach to planning and management of 
mangroves, the forest management plan should conform to an approved 
Integrated Coastal Area Management programme. 


(a) MAI22M: In support for the above policy, coordination among 
concerned agenciea and land ueere is required. To coordinate mnd 
promote environmentally sound development of mangrove resources, 
a National Mangrove Coordination Committee should Jbe established. 
Membership of the NATCOM comprises concerned Ministries and 
Departments, Research and Education institutions and HQOs. 

(b) Consensus: To achieve a national consensus, seminars and 
workshops to discuss mangrove management policy and land use 
should be organized by NATCOM. 

(c) Development strategy: A strategy for the sustainable management 
and use of land and aguatic resources, Including conservation of 
wildlife and biodiversity, should be prepared and implemented* 

(e) Coordination: Close liaison should be fostered Jbetween the 
Services of Forestry, Fishery, Environment and Agriculture, 
related departments and NGOs, who have an interest in the 
management, utilization, regulation or research in mangroves. 

(d) Social forestry: In forestry, there is an economic and a welfare 
sub-sector. Meeting the basic needs of the poor is mainly the 
responsibility of the public sector. Sufficient mangrove forests 
should be designated for the local supply of goods and services 
needed by the rural communities. 

15.1.3 Legislation 

Generally, the coastal zone is covered by several overlapping 
maritime, fishery, forestry and other laws. It is important that the 
agreed policy should enjoy the force of law, which does not conflict 
with the existing laws of the country. 

Legislation on conservation of mangroves should be framed, 
revised or amended to reflect the agreed forest policy. Such a law 
should harmonize with the general body of legislations in the country. 


The nature, form, and extent of mangroves should be determined through 
national inventories: 

The mangrove resources (i.e. terrestrial and aquatic), regardless of 
ownership status, should be inventoried to assess their relative economic and 
ecological importance and management requirements at national and local 


A permanent resource base forms the basis for sustainable management and 
use, in order to optimize their contribution to national development. 

To ensure protection and conservation, mangroves, regardless of their 
ownership, should be constituted as mangrove reserves, and sufficient areas 
be reserved for production purposes. 


15.3.1 Classlf ication of forest use categories 

The economic potential of the mangroves stems from three main 
sources: viz., forest products, marine products and e cot our ism 
(production function) . In addition, mangroves have protection 
functions. There are also marginal mangroves in the inter-terrestrial 
that are more suitable for permanent agriculture or other non-wood 
uses, (conversion forests) . 

Mangroves should Jbe classified into three functional categories 
as follows: (a) Production; (b) Protection, and (c) Conversion forests. 
The last category may be converted to other uses subject to 15.1.1 (a) . 

15.3.2 Land tenure and usufruct 

The customary and usufructuary rights of indigenous people and 
rural communities over forest produce should be clarified as part of 
the enabling mechanism devised to promote broad-based participatory 
forestry programmes. 

To augment tJbe resource Jbase, community-based mangrove 
plantations and private woodlots should be promoted, particularly in 
fuel wood deficit areas. Concomitant measures to clarify the 
usufructuary rights and land tenure arrangements for rural communities 
should be made. 


Establishing a mangrove management unit within the Forest Service, that 
is responsible for management planning, harvesting, reforestation and 
protection is highly desirable in countries endowed with abundant mangrove 
resources . 

15.4.1 International technical assistance 

In countries where the management, protection and integrated use 
of mangrove resources have not been institutionalized, assistance in 
strengthening the technical and managerial capacity and capability of 
the concerned Forest Services is required. 

International assistance may comprise the following: 

Fellowships and study tours; 

' Build up a working library on mangrove ecosystems, management, 
utilization, wildlife conservation and protection; 

m Organization of seminars, workshops and training courses in 
academic, research institutions and NGOs; 

Pilot demonstrations. 

Provide international expertise in preparing model management 
plans for selected pilot areas and to provide on-hand experience; 

Provide technical assistance to formulate and implement 
appropriate research programmes. 

Adequate funding and vehicles and boats, equipment, materials and 
tools to undertake sustainable mangrove management should be provided. 
Back-up research support is also needed. 


The continued production of wood and non-wood benefits is greatly 
dependent upon the effectiveness of forest management measures. 

(a) Management objectives! Management objectives, which meet the 
socioeconomic, technical and environmental requirements of the 
forests should be framed as a basis for action. (See Chapttr 9.) 

(b) Management planning! The first vital requirement for forest 
management per se is the preparation of long- term management 
plans, integrating the production of selected wood and non-wood 
resources, needs of rural population, recreation, conservation of 
genetic resources and soils/water protection. 

(c) Growth and yield plots: Scientific management depends on, inter 
alia, growth and yield data, and information on the regeneration 
and phenological characteristics of desirable species. 
Accordingly, growth/ yield and silvi cultural plots should be 
established to provide growth and biological data for forest 
yield prediction, regulation and management. 

(d) Silvicultural concept: The silvicultural concept should aim at 
the cost-effective sustained yield of desired products without 
impairing the environment. The choice of silvicultural systems 
should take into account the requirements of the multiple-use 
concept elaborated under 15.1.2. 

A clear -fell ing system with the retention of standards using the 
alternate felling strip method has been successively used in many 
countries. The rotation age may be 9-12 years for firewood or 
stands managed under the Coppice-with-Standards system or 25-35 
years for vigourous Rhizophora stands. (Refer to Chapter 10.) 

A selection felling system based on a minimum cutting diameter 
(say, 15-18 cm overJbarJU should Jbe applied where the overriding 
management objective is to optimize non-wood production or 
conservation (e.g. mollusca culture, breeding or feeding of 
commercial shrimp and fish species, or wildlife). 

(e) Annual Allowable Cut; The annual allowable cut for wood 
products, permissible fishery catch, and use levels for services 

(ecotourism) should be set flexibly, conservatively and 
harmonized to ensure sustainable management. (Refer to 
Chapter 11.) 

(f) Harvesting Plan; Harvesting must be regulated by a harvesting 
plan. The harvesting system used must be compatible with 
ecological and environmental site requirements apart from being 
cost -effective. It should also not impair the production of 
other non-wood resources. 

Post harvest surveys are necessary to assess logging damage, 
regeneration sufficiency, and to plan follow-up treatment 
measures required. 



Equitable distribution of forest management incentives, -costs and 
benefits between the forest authority, forest owners, rural communities and 
private entrepreneurs is a vital requirement. 


(a) "Voice of the people 11 ; Management plans, no matter how well 
articulated, cannot succeed without regard to the requirements 
and aspirations of indigenous and local populations. Success 
depends, inter alia, on being able to match management objects 
with the interests of local populations and through extension, 
secure their support and commitment, 

(b) Local participation; Participation of rural communities in 
mangrove -based small-scale industrial activities should be given 
priority. It is also most desirable to encourage a degree of 
"self -management 11 amongst the various users of the mangrove 
environment, such as shrimp- farmers, farmers, fishermen, 
charcoalers so that they may be involved in protecting their own 
resources . 

(c) Sensitization: The potential of mangroves for rural development 
is not perceived. The principal issue, therefore, is to ensure 
that planners and decision makers have access to factual 
information on the role and potential of mangroves. Seminars, 
talks, workshops, film shows and exhibitions on mangrove products 
and services targeted for various audiences should be conducted 
to create public awareness. 

(d) Providing economic alternatives: Over-exploitation occurs due 
less to a lack of awareness of the problems, than to a lack of 
economic alternatives. To win public acceptance and support for 
forestry programmes, in situ pilot projects are required to 
demonstrate the economic viability, sustainability and 
manageability of planting mangroves, proper carbonization and 
sound forest management practices, as well as ecotourism; 

(e) Staff training: Professional forest officers should attend short 
courses on mangrove ecology, conducted by the local universities 
wherever possible, to widen their knowledge and appreciation of 
the technical basis needed for successful integrated forest 
management ; 


Rational management is based on an in-depth understanding of the forest 
and its environment that can only be obtained through a series of planned 
observations and measurements relating to its composition, structure and 


15.8.1 Environmental Impact research 

Multidisciplinary studies to determine the biological, physical 
and socioeconomic effects of the major users of mangrove areas is 
required. The objectives are to determine: 

the inter-relationship between maricuiture, current forestry 
practices and other human activities and their resources/ 

the costs and benefits, both social and economic, of different 
alternative uses; 

alternative aguaculture systems consistent with local ecologies; 

criteria for site selection of mangrove areas for aguaculture 

15.8.2 Socioeconomic studies 

The objectives are to analyze: 

the social and economic setting of the mangrove dependent 

income levels, income sources, income distribution in such 
coastal communities; 

alternative economic activities and investments; 

market structure for mangrove products; and 

practices /techniques used in collection of mangrove products. 

15.8.3 Demand for mangrove products 
Objectives are: 

to estimate the local, regional and/or national demand for 
mangrove products, including their export potential; 

to analyze current and future supply and demand trends; 

to determine price trends; and 

to analyze factors affecting the demand for such products. 

15.8.4 Evaluation of mangrove management policies/programmes 

The objectives are: 

to review and formulate policies on the utilization, management, 
conservation and wildlife and research; 

to determine the extent of implementation of such policies and 
programmes; and 

m to analyze the impact of such policies and programmes on the 
mangrove -dependent population. 

15.8.5 Ecological and silvicultural studies 

Studies on plant succession, structural development, phenology, 
effects of treatment and other related aspects should be initiated as 
the findings are highly relevant to the formulation of appropriate 
silvicultural systems and management prescriptions. 

15.8.6 Zonation and Bite classification 

A Study on the relationship between vegetational zonation, tidal 
influences, salinity and soil conditions is required to determine the 
most appropriate silvicultural treatment to apply, species to be used 
in reforestation or afforestation, site potential and also integrated 
land use planning. 


(a) Contamination: Mangroves are influenced by tidal as well as 
terrestrial influences, in terms of their hydric, biochemical and 
ecological impacts. Aquatic animals and terrestrial plants are 
affected by water quality and sedimentation. Monitoring of industrial 
and man-made pollutants that may adversely impact on the mangrove 
ecosystem should Jbe conducted and regulatory measures taken to minimize 
estuarine and coastal contamination. 

(b) Upland land practices; Unregulated river flows and heavy siltation due 
to land mismanagement in the watersheds and the extensive use of 
pesticides are producing negative impacts in the mangroves. 

(c) Wildlife management ; Mangroves provide habitats for a variety of 
wildlife and avifauna. Surveys on populations, feeding and migration 
of major wildlife species should be conducted. These data could be 
used to formulate a plan for the interactive management and control of 
these species. Through proper planning and management, the wildlife 
potential can not only be preserved but also commercially exploited and 
the recreational potential of the area enhanced. 

(e) Crocodile farming; The economic and technical feasibility of crocodile 
framing should be explored, where there are sufficient wild populations 
and the supporting facilities are adequate, including feedstock. (Refer 
to Chapttr 3.) 

(f) Apiculture: Most mangroves have the potential to support a thriving 
honeybee -keep ing industry which is environmentally compatible. A 
survey should be initiated to determine the extent of melliferous 
plants (honey and pollen resource) with the view towards improving or 
introducing local apiculture as an in -situ income source to coastal 
mangrove dwellers. (See Chapter 3) 

(g) Scot our ism: if properly implemented ecotourism can provide people with 
a viable alternative to destroying their environment. 

The mangroves have many natural scenic, vegetational features, fishery, 
wildlife and birds which are attractive to ecotourists. The potential 
for ecotourism as a viable economic alternative to rural people and to 
conserve the environment and its natural resources should be fully 
explored. (See Chapter 3) 






1.1 Objectives 217 

1.2 Aerial photo- interpretation 217 

1.3 Analysis of Landsat digital data 217 

1.4 Results 218 


2.1 Forest type classification 221 

2.2 Forest inventory 222 



3.1 Stratification 224 

3.2 Forest type classification 224 

3.3 Sampling design and field measurements 224 

3.4 Volume determination 225 

3.5 Results 225 



4.1 Introduction 227 

4.2 Methods and materials 227 

4.3 Results 227 

4.4 Discussion 228 




5.1 Introduction 229 

5.2 Metods and material '. 229 

5.3 Results 233 



6.1 Introduction 235 

6.2 The flora of the Sundarbans 237 

6.3 Assessment of the wood resources 238 

6.4 Fauna of the Sundarbans 240 

6.5 The socio-economic importance of the Sundarbans 241 

6.6 Management of the forest - past and present 243 

6.7 Present management administration 245 

6.8 Sustainable management of the wood resources 246 

6.9 Management of minor wood and non-wood products 252 

6.10 Wildlife management 254 

6.11 Conclusion 254 

Bibliography 256 




l.l Comparison of classification results in Klung District, 

Thailand in the 10 year period 1975-1986 218 

4.1 Inventory results, Ma-swar, Sierra Leone 228 

6.1 Summary of net merchantable volumes in the 1950s 239 

6.2 Area of major forest types 239 

6.3 Merchantable volume of different species in the 1980s 240 

6.4 Annual production from the Sundarbans 241 

6.5 Ranges in the Sundarbans 245 


1.1 Map based on photo- interpretation of aerial photography .... 220 

1.2 Map based on digital analysis of Landsat image 220 

2.1 Forest type mapping on stereo aerial photography 222 

2.2 A mangrove forest type map based on interpretation of 

aerial photos ' 223 

3.1 Part of the forest type map of the Sundarbans 226 

5.1 Mature natural mangrove forest 231 

5.2 Measuring the felled tree 231 

5.3 Weighing the stiltroots of a felled tree 232 

5.4 Measuring the green density of wood samples 232 

6.1 Map showing the Sundarban Mangroves 236 

6.2 Bundling of Excoecaria wood 251 

6.3 Excoecaria wood ready to be transported to the Newsprint Mill . 251 

6.4 A boat loaded with Nypa leaves 253 

6.5 A fishing boat and crew 253 

6.6 Spotted deer coming down to drink at a man made water hole . . 255 

6.7 The Sundarbans - a mangrove forest with great biodiversity . . 255 


l.l A flowchart of the digital image analysis 219 


6.1 General timber harvesting rules 247 

6.2 Timber harvesting rules used by the Newsprint Mill 249 

6.3 Timber harvesting rules used by the Match Factories 250 

6.4 Harvesting rules for Nypa (Golpatta) fronds 252 



(Source: Ratanasernmong, 1986) 


One of the objectivee of this survey was to compile information on 
mangrove distribution and classification and to produce thematic maps. On a 
few test sites, both visual photo-interpretation and digital analysis of 
Landsat data were conducted to enable a comparison of the two methods* 


Black and white photos at a scale of 1:15 000 scale taken in 1975 and 
infra-red photos in the scale of 1:20 000 from 1985, were interpreted for the 
purpose of identifying and assigning "training areas" for satellite data 
analysis. Photo- interpret at ion permitted the differentiation between the 
following nine classes: 

1. Mangrove forest class I: Located near the sea shore, on a soft and wet 
soil. The forest is dominant ly composed by RhJLzophora mucronata and 
Bruguiera gymnorrhiza. In other areas, the dominant species include 
Rhizophora apiculata, Bruguiera gymnorrhiza, Avicennia sp. and 
Xylocarpua moluccensis; 

2. Mangrove forest class II: Located on lower land than class I. This zone 
remains submerged after the high tide. It is mainly composed by pure 
stands of Ceriops tagal. In the transition area between class I and 
class II, Acrosticum aureum and Xylocarpua ap. are present; 

3. Mangrove forest class III: This class is located in higher land, on 
hard soil, which is flooded only occasionally. The main species are 
Lumnitzera littorea, Lumnitzera racemosa, Phoenix paludoaa, Intaia 
bijuga and Melaleuca sp.) 

4. Mangrove plantations; 

5. Mangrove clearings (for shrimp farms); 

6. Paddy land; 

7. Shrimp farms; 

8. Orchards and other solitary standing trees; 

9. Water. 


Images from Landsat bands 4, 5 and 7 were used in the analysis. Image 
enhancement was also carried out by using a band ratio of difference over sum 
of bands 5 and 7 (5-7/5+7). 


The classification was based on the training areas derived from the 
results of photo* interpretation. The maximum likelihood approach was applied 
in the image classification. A filtering process was further applied in order 
to reduce the patchy appearance of the classified features. Diagram 1.1 
illustrates the sequence of the image analysis process. 


Figures 1.1 and 1.2 are the outputs obtained from photo -interpretation 
and digital analysis of the Landsat image respectively. A summary of area 
statistics is presented in fable 1.1 below. 

Table 1.1: Comparison of classification results in Khlung District, Thailand, in the 
10 year period 1975-1986. 


Air photos Landsat IR photos Landsat 
1975 Nov. 1975 Oct. 1985 Jan. 1986 
km2 km2 km2 Km2 







Mangrove class I 
Mangrove class II 
Mangrove class II 


Mangrove class III 
Mangrove plantation 
Clearing area 
Shrimp farms 
Paddy fields 
Standing trees 

As can be seen from the table above, it was found that the confusion 
between the 3 mangrove classes is rather high, suggesting that these 3 classes 
should be grouped as one . 

The accuracy of the classification could likewise be improved 
considerably if classes 5 and 7 (Mangrove clearings for shrimp farms, and 
Shrimp farms) were to be grouped together. 









Band Ratioing 
(Vegetation index 

Training area 


Maximum likelihood 



I Report of 

/ result 


'Colour Map 

^ Mangroves 

(Source, ..port on t. Sensins .* 

"!. TWl-d. > 



ffi&fl Mangrove forest class 1 

gw| Mangrove forest class 2 

P*3 Mangrove forest class 3 

E23 Paddy land 

Bgl Shrimp farm 

(*:::< Standing trees 

R5?3 Mangrove clearing 

f I Water 

Fig. LI: Map based on photo-interpretation of m aerial photography 

Khlung District, Chantaburi, Thailand 

mmm Mangrove! 

_ Mangrove II 
***** Mangrove HI 
Mangrove clearing 

Shrimp farm 
Paddy land 
standing trees 

Flf. U: 

MSS bands 4, 5 and 7 used. Same area as above. 




(Source: Luna Lugo, 1976) 

The total area concerned by the study encompasses about 20,000 ha, 
including water and mangroves without commercial value. 


On 1:25 000 scale panchromatic aerial photographs, three mangrove genera 
were distinguished, based on tone, relative height and crown diameter -as it 
is presented in the following table. 









A further subdivision was achieved using tree height and canopy density. 
Height and density classes adopted are listed in the two tables below. 

Height classes: 


Total height 

Type 1 


> 30 m 
20-30 m 
10-20 m 
< 10 m 

High 1 
Very low 

Density classes; 


Density (%) 

Canopy cover 



Very dense 




















The height was occasionally measured with a parallax bar, and density 
was estimated using a transparent grid. 

Photo- interpretation at that scale permitted the establishment of the 
spatial distribution and degree of mixing of the three genera of mangrove, but 
did not allow the distinction between the three species of Rhizophora present 
in the area, namely R. mangle, R. racemosa and R. harrisonii. 


A forest inventory was carried out using a systematic sampling design 
with 1% intensity. Strips, 20m wide and 300m long were laid out 
perpendicularly to rivers and channels. The distance between strips was 
2 000m, 

In the strips, all trees with dbh larger than 8cm were measured and 
their species identified. Also, for the regeneration assessment, seedlings 
were counted in 2mx2m plots located along the strips. 

With regard to timber volume calculation, tables for volume, bark and 
form factor were constructed for each diameter class. Moreover, from 
interpreted aeriSw photos (of which an example is presented in Figure 2.1) a 
forest type map was produced. Figure 2.2 is an example of such a map. 

A : Rhizophora 
B : Avio*nnit 
C : Marsh land 

Fig. 2.1: Forest type mapping on stereo aerial photography 

Vencrudt. (Source: Luna Lugo, 1976.) 


dc Por.q 

Fig. 2.2: A mangrove forest type map based on interpretation of aerial 

Venezuela. (Source: Luna Lugo, 1976) 



(Source: ODA, 1985) . 

The Sundarbans forest is the largest single-tract mangrove forest in the 
world (4 200 km a in Bangladesh alone) and has been managed for more than a 
century. Currently, it is a major source of timber, fuel-wood, pulp wood and 
various other products. The main objectives of the inventory which was 
carried out in 1983 and 1984, were to produce forest type maps and to assess 
the growing stock. 


The area was divided in nine blocks. The demarcation of the blocks was 
intended to reflect broad differences in forest types. In each block, 
preliminary air photo- interpretation was carried out on panchromatic prints 
of the scale 1:30 000. 


Forest types were subdivided into canopy density classes. Moreover, the 
main species, namely Heritiera foznes and Excoecaria agallocha were subdivided 
into height classes by inspection of the height data collected during 
enumeration. Density and height classes were as follows: 


Stand density 


Height (h) 


> 70 % 
30 - 70 % 
10 - 30 % 


h >15 m 
15 m > h > 10 m 
10 m > h > 5m 
h < 5 m 


Each sampling block was sampled individually. In those blocks 
containing the most important species, the sampling design was based on forest 
type and compartment. Approximately ten sample plots per compartment were 
randomly located in each type accounting for more than 10% or more of a 
compartment area. In the case of forest types not occupying as much as 10% 
of any compartment area but occupying 5% or more of the total block area, 
sample plots were randomly located through the block as a whole. Table 3.1 
illustrates the various sampling units used in each block. 

In each plot, dbh measurements were taken on all trees with a diameter 
equal to or larger than 5 cm. The height of the tallest trees in the vicinity 
were also measured. 


Regeneration was assessed by counting the number of small trees having 
a dbh less than 5 cm and those below 1.30 m. Other data on the soil and on 
the top dying of Heritiera fomes were collected in field sample plots. 

For growth determination of the main species, ring counts were made on 
disks collected from felled trees, but information obtained from periodic 
measurements from existing permanent plots was also used. 


Volume regression equations were established for the main species 
mentioned above, based on measurements taken on felled trees used in increment 


3 . 5 RESULTS 

The results of the inventory include among other things: 

Areas of forest types, by canopy cover and height; 

Areas by compartment for all forest and non forest categories; 

Volume tables and regressions for all the main mangrove species 
with both stem volume and crownwood volume; 

Volume of the growing stock and bark for each commercial timber 
species, by forest type and block; 

Diameter increment by height classes for the main species. 

Some of these results can be found in Cat* Study 6. Examples of the 
volume regressions obtained are presented in Appendix 5. 

Type maps were also produced. An example is shown in Figure 3.1. 












in in 


Q P 

2 m in 

- S n 
jo o ,4 

in ol 



At V 



I A 

> V 


^ V 



' 1 I s s 


< 5 


i ' < 


*- Q 

S 8 



I i I i 


d b 



o o do 




o c> 

So * 

o o o 

(M 01 01 






X 3 


i i 


s | 1 

^ in 01 X 









L | 








A ? 






> I 



1 i 

: : 

9) 0) 


w 10 c 







I | 
S c 






I J fi 



































y 1 












in iji 











^^^^ _-^ 



, y 
































Table 3.1: 

Types of samples units, steas recorded and 

blocks In which used. Survey of the Sunderbans. 



SCALE : 1:50 000 
(The code* indicate the foreet types defined above.) 

Fig. 3.1: Part of the forest type map of tteSundariwuis 

Bangladesh. (Source: ODA, 1985) 




(Source: B.Birkenhager, 1988) 


Under the FAO/UNDP Fuelwood Project in Sierra Leone (SIL/83/003 and 
SIL/88/008), a pilot area of 270 ha was selected for demonstrating improved 
mangrove management techniques. 

The area, on three sides enclosed by a U-shaped curve in the Ribi River, 
is inundated by normal high tides and consists of an almost intact natural 
mangrove forest. The main species encountered were Rhizophora racemosa, which 
formed the highest stands - especially along the river; Rhizophora harrisonii 
and Avicennia africana. 

The forest inventory was undertaken in 1987/88. 


The systematic line plot sampling design was employed with 100 m between 
the lines and 40 m between each circular plot along the line. The radius of 
the plots was 5 m giving a plot size of 79 m 3 and a sampling intensity of 2 %. 

Since initial reconnaissance and the study of aerial photographs 
(1975/76 Infrared False Colour; scale 1:70 000) indicated the occurrence of 
strongly divergent vegetation types within the area, the vegetation type of 
each plot was noted based on the following stratification: 

(a) High Forest (Pure Rhizophora spp.) 

(b) High Forest (Rhizophora/Avicennia) 

(c) Semi-high Forest (Rhizophora) 

(d) High Bush (Pure Rhizophora spp.) 

(e) High Bush (Rhizophora/Avicennia) 

(f) Low Bush (Pure Rhizophora spp.) 

(g) Low Bush (Rhizophora/Avicennia) 

Within each plot the species and the dbh for all trees with a dbh * 7 
cm were recorded. (In the case of trees with stilt roots originating above 
1 m up the stem, the diameter 30 cm above the stilt root was used.) The 
height of trees was measured on all trees in every fifth plot. 

A total of 655 plots were laid out. With 8 plots falling in the Ma-swar 
River, the assessment was based on an inventory of 647 plots. 

4 . 3 RESULTS 

Based on the classification of plots into vegetation types and with 
additional information obtained from infra-red aerial photographs at a scale 
of 1:70 000, a vegetation map was produced. The map clearly shows that the 
high forest is found along the river banks and that towards the interior of 
the swamp a decrease in height occurs via semi -high forest and high bush to 
low bush. 


Matching plot data and plot stratification led to the following 
characteristics for the vegetation types discerned in this study: 

Table 4.1: Inventory results, Ma-swar, Sierra Leone 






timated volume (m'/ha) 
Mean volume 90% confidence limits 
Rhizophora Avicennia Rhizophora Avicennia | 

High Fortftt 






High Pormmt 








Smi -high 




59. f 


High Buah 







High Bush 








Low Buh 






Low Bush 

l _ B _i^ MI _ MMIB __ B ___ Ma ___ ll __ 








The volume in the above table is the stem volume above 7 cm and is 
estimated using a volume table from Matang, Malaysia with entry for dbh. 

The total standing volume of the area was estimated to be 23 000 m\ 
With a suggested rotation of 15 years, a maximum of 1 500 m 3 can thus be 
harvested annually during the first rotation if a clearfelling system is to 
be employed. 


The systematic line plot sampling design and the sampling intensity 
employed yielded adequate results. However, the actual plot size was deemed 
to be too small as only 2 plots contained 10 trees or more. This accounts for 
some of the variance of the estimates - especially with regard to plots found 
in the mixed Rhizophora/Avicennia vegetation types, as the actual plots often 
contained either Rhizophora or Avicennia trees rather than a mixture of the 

For natural stands such as these, comprising mainly Rhizophora spp., 
where in many cases the stands are far from being homogenous and the density 
of trees is small due to wide spreading stilt roots, it is thus recommended 
that a plot radius of 8 or 10 m be used. 

The volume table used was the only one available at the time, but is 
based on other species (Rhizophora apiculata and Rhizophora mucronata) grown 
as even-aged plantations, not natural stands. It was thus decided that a 
local volume should be constructed to test the validity of the Malaysian table 
and, if necessary, to adjust the above estimates. See Cms* Study 5. 



(Source: M.L0yche and C.L. Amadou, 1989) 


As a follow-up to the inventory carried out as described in the previous 
case study, a local volume table was constructed due to the following reasons; 

No volume table was available for the mangrove species of West Africa; 

The volume table from Matang, Malaysia used for the preliminary 
assessment of wood volume in the Ma -s war area in Sierra Leone is based 
on Rhizophora apiculata and Rhizophora mucronata grown as even aged 
plantations and only gives the big wood volume i.e. stem volume above 
7 cm in diameter with the entry for dbh; 

In Sierra Leone, the mangroves are mainly natural and mature climax 
stands, which along major rivers reach a height of up to 40 metres. 
The dominant species is Rhizophora racemosa, which often have stilt 
roots originating 4-6 m up the stem and dbh is thus not a valid 

Near urban centres, in particular close to the capital Freetown, over 
exploitation by unregulated cutting for firewood and charcoal has 
reduced former mangrove forests to thickets seldom reaching above 5-6 m 
in height and 10 cm dbh. The main species here are R. racemosa , R. 
harrisonii, R. mangle and Avicennia africana. Puelwood is still being 
extracted here utilizing wood down to approximately 3 cm in diameter. 

The volume table should thus take into account not only the big wood 
volume, but also indicate the wood available from the above over exploited 


The area chosen for felling and measuring trees was the Ma-swar mangrove 
area, where the inventory described in Case Study 4 had taken place. 

The preparation of a volume table normally entails the felling of a 
great number of trees. However, due to limited skilled manpower, equipment 
and funds, it was decided to measure approximately 100 trees equally spread 
over the area to avoid 'neighbouring effects' and in order to have all the 
vegetation types represented. At the same time it was decided to obtain a 
maximum of 20 measurements for trees with a dbh less than 10 cm and 15 trees 
in each of the following 10 cm diameter classes. 

The outline of the inventory plots (the line plot sampling design) was 
followed and it was decided to fell one tree in every fifth plot. Thus the 
distance between the plots along the line was 200 in with 100 tn between each 
line of plots. 


The tree to be felled in each plot was chosen at random. An attempt was 
made to chose a tree with a diameter close to the average diameter in the 
plot. Abnormal trees with broken stems, dead tops etc. were avoided, as were 
trees directly bordering the main river. Plots only containing trees with 
diameters falling within a diameter class, where the maximum number of trees 
had already been reached, were disregarded. 

After choosing the tree, the diameter was measured with a diameter tape. 
This was in most cases the 'diameter above stilt roots 1 (D M ) , taken 30 cm 
above the highest stiltroot, as recommended by FAO (1980) . Only stilt roots 
resulting in a deformation of the stem were taken into account. 

The stem was marked with lumber crayon at the measuring point and the 
height above ground level of this point recorded and the tree was then felled 
with a chainsaw. 

Starting from the point of the D a , the stem was then marked for each 
one meter following the form of the tree, and the diameter at the end of each 
section recorded in cm with one decimal, until a diameter just above 7 cm was 
reached. The exact point where the diamter was 7.0 cm was found and marked, 
and the distance to the last recorded diamter measured. The point where the 
diameter was exactly 5.0 cm was likewise marked and the distance to the 7 cm 
mark recorded. 

The length of the tree from the D M mark to the top most leaf bud 
following a straight line was recorded and added to the height at which the 
D M was measured to obtain the total height of the tree. 

All branches and stilt roots down to 5 cm in diameter were weighed. As 
the branches often were very crooked and the stilt roots oddly shaped, this 
was found to be the best method of obtaining an estimate of the amount of this 
wood, which in many cases would be used as firewood. 

For trees with a dbh less than 10 cm, a slightly different method was 
employed, as these trees are often felled for firewood by the local villagers 
using cutlasses, and wood down to 3 cm in diameter is extracted. 

The diameter and the total height were recorded as previously, but 
instead of measuring the volume, the tree with branches and stilt roots down 
to 3 cm in diameter was weighed. 

Establishment of the green weight and volume of samples from some of 
these trees and of some of the branches and stilt roots of the bigger trees 
was undertaken in order to determine a conversion factor from weight to 
volume. The volume of the wood samples was found by submerging them in water 
and measuring the weight and thus the volume of the water displaced. 

A total of 85 trees were felled and measured. Three wood samples taken 
from each of 10 trees with a diameter below 10 cm dbh; 30 samples of branch 
wood and 30 samples of wood from stilt roots were weighed and their volume 
measured to obtain the green density. 

Please refer to Figures 5.1 - 5.4 for an illustration of the difficult 
working conditions found in mangrove areas. 


Figure S.I: Mature natural mangrove forest, Ma-swar, Sierra Leone 

Photo by M.L.Wilkie 

Figure 5.2: Measuring the felled tree, Mawar, Sara Leone 

Photo by M.L.Wilkie 


Figure 5 J: Weighing the sttttroots of a felled tree, Ma-swar, Sierra Leone 

Photo by M.L.Wilkie 

figure 5.4: 

MeMurtnf the 

Photo by M.L.Wilkie 

Ma-war, Sierra Leone 



For the volume calculations the Smalian formula 

V ......... x - x L 

2 4 

was used to determine the volume for each section of the stem above D... 

Due to the shape of the stem below the stilt roots, this part was 
regarded as part of the stilt root system. 

For the trees with a diameter above 10 cm, the stem volume above stilt 
roots and above 7 cm in diameter was calculated, as was the stem volume down 
to 5 cm. The results from the determination of the green density of branch 
wood and stilt roots were used to convert the measurements of weight into 
volume and the total wood volume above ground down to 5 cm. in diameter was 

For trees less than 10 cm in dbh, the green density obtained was used 
to convert tho weight into volume and the total volume down to 3 cm in 
diameter was calculated. 

The results of the determination of the green density are shown in 
Table 5.1 below: 

Table 5.1: Results of the determination of the green density 

of wood 

No, of 

(g/cm 1 ) 


95% confidence 




















IHM ,^^__^^____ MMHHM _ HM _i 

As the wood from the small trees was slightly heavier than the branch 
wood and significantly heavier than the wood from the stilt roots, the 
individual means were used for the conversion rather than an overall average. 

Plotting of the diameter against the height of the trees measured showed 
a very good correlation in spite of localised variation in site factors and 
the use of D M for most of the bigger trees . The correlation between diameter 
and height was analysed and curve fitting with four different regressions 
(Logarithmic curve, Straight line, Exponential curve and Power curve) was 
undertaken. The heighest correlation coefficient (0.9960) was obtained for 
the power curve derived from the following formula: 

H - 1.123 x D- 1592 


where H is the total tree height in metres and D is the diameter in cm 
measured at breast height (1.30m above ground level) or - especially on the 
bigger trees - 30 cm above stilt roots. 

Due to the good correlation between diameter and height, it was decided 
that a one entry tariff for the volume table would be sufficient. An analysis 
of the relation between the diameter and the volume was thus undertaken in the 
same manner as above. For the relation between diameter and stem volume above 
7 cm, the best correlation coefficient (0.9986) was obtained for the following 

. 0.0001 x D 2 -* 471 

where V 9tM is the stem volume in m 3 above the point of the dbh/D M and 
down to 7 cm in diameter and D is measured as above. Based on the above 
formula a volume table for the Rhizophora species in Sierra Leone has been 
constructed. This table was found to vary considerably from the' Malaysian 
volume table previously employed (See Appendix 5) . 

The correlation between diameter and the total tree volume was, as could 
be expected, less good. However, a correlation coefficient of 0.9942 was 
found for the following regression: 

0.004 x D 2 2128 

where V total is the total tree volume in m 3 down to 5 cm in diameter. 

For trees less than 10 cm in dbh, the total tree volume down to 3 cm in 
diameter was found to average 0.0199 m 3 . 

A summary of the findings with regard to trees above 10 cm in dbh is 
presented in Table 5.2 below: 

Table 5.2: 

Summary of flndtafs according to diameter classes 


No. of 



Avenge V^ above 7 cm 

Avenge V^ above 5 cm 





Actual Regression Diff. 

Actual Regression Diff, 





(m*) (m 1 ) (%) 

(m*) (m) (%) 1 











! -12.2 1 










11.1 1 




















- 3.7 .] 














M __ BHHB ^__ M I 







A0 the number of trees measured is relatively small, the above volume 
regressions are only tentative in nature, but serve as a good starting point. 
More data should be collected from a variety of sites within the country to 
further refine the above table (s). 




(Source: M.Z. Hussain, 1992) 


The Bangladesh portion of the Sundarbans mangrove forest is located in 
the Southern most extremity of the Qangetic delta, bordering the Indian State 
of West Bengal in the west and the Bay of Bengal in the south. The forest 
occupies a flat mud swamp which is submerged by high spring tides most of the 
year and by almost all high tides during the rainy season. 

This portion of the Sundarbans covers an area of 577 285.6 hectares. 
The forest extends into the State of West Bengal where it covers an additional 
area of 416 000 hectares. Within Bangladesh, the Sundarbans forest extends 
over three administrative districts in the Southwestern part of the country 
and lies between 89B and 90E longitudes and 21ON and 2231N latitudes 
extending about 80 kilometers inland from the Bay of Bengal. Please refer to 
Figure 1. 

The total land area of the Bangladesh portion of the Sundarbans is about 
401 600 hectares. The rest is under water in the form of rivers, channels and 
creeks varying in width from a few feet to several miles. The larger of these 
rivers are the remains of former beds of the Ganges, which have gradually 
shifted eastward, and are no longer directly connected to the Ganges river. 
Baleswar, which passes along the eastern boundary of the forest, is the only 
river which is still connected with the Ganges and receives direct fresh water 
effluence from this river. A large number of channels and creeks flow into 
larger rivers in the Sundarbans. These, in addition to flooding the forest 
floor, make most of the forest accessible by country boats during high tides 
and make forest extraction activities relatively easy. Some of these creeks 
or channels flow between two major water ways and play important roles in 
navigation within the forest. 

Tides also play an important role in the Sundarbans since the scouring 
action during the ebb tides remove silt that may have been deposited and keep 
the rivers, channels and creeks open. The high tides on the other hand ensure 
the inundation of the forest floor - a necessity for the proper growth and 
sustenance of the vegetation. 

The Sundarbans has a humid tropical climate with a mean annual rainfall 
ranging between 1 640 and 2 000 mm. The quantity of annual rainfall recedes 
from east to west. Most of the rainfall occur between May and September. 
There is very little rainfall between November and April. The temperature 
ranges between 20.4C and 31.5C. The highest temperatures are recorded 
during May and June and the lowest during December and January. 





100 Km 

Divisional Bdy 









D H A K 









23* n 


Mp showing the SundaitanMtnfroves 


The forest grows on soil which is of recent geological origin and 
consists of alluvium deposit which has been washed down from the Himalayas. 
The soil is a silty clay loam and the sub-soil consists of alternate layers 
of clay and sand which at greater depths are compacted with shale and 
sandstone. No rock formation occurs in the area. The surface soil of the 
forest area consists mainly of clay. Sandy patches are absent, except near 
the sea. The grey colored soil is neutral to mildly alkaline in reaction. 
However, in some areas the dried up sub-soil is more acidic in reaction. The 
soil is slightly saline in the East/Northeast and moderately saline in the 
West /Southwest. A seasonal gradient of salinity fluctuation in soil and water 
exists in the Sundarbans. The salinity gradient reaches a peak in April - May 
before the rains but falls off abruptly with the monsoon rains in May - June. 


, Heritiera fotnes (Sundri) and Excoecaria agallocha (Gewa) are the two 
principal mangrove tree species in the Sundarbans. According to a recent 
inventory report (Chaffey, 1985), Sundri and Gewa occur over more than 
70% of the Sundarbans in either pure patches or mixtures. One of the two 
species occurs in mixture with other species in over another 25.8% of the 
forest. Ceriops candelleana (Goran) is the third most frequent species 
occurring in the Sundarbans. Other important tree species in the Sundarbans 
are Sonneratia apetala (Keora) , Xylocarpus mekongensis (Passur) , Bruguiera 
gymnorrhiza (Kankra) , Avicennia officinalis (Baen) , Cynometra ramiflora 
(Singra) , Amoora Cucullata (Amur) , Hibiscus tiliaceus (Bhola) and Nypa 
fruticans (Golpata) . A complete list of the Sundarban flora can be found in 
Chaffey et al. (1985) and is also included in Das and Siddiqi (1985). 

The Eastern portion of the forest receives much more fresh water than 
the Western portion and the quality of crop is better in this region due to 
the fresh supply of silt each year, which result in. a soft and fertile top 
soil. The quality of crops gradually deteriorates towards the West, where, 
in the absence of silt, the top soil has become hard and less suitable for 
tree growth (Das and Siddiqi, 1985) . 

Heritiera fomes prefers relatively fresh water area and is thus more 
abundant in the Eastern portion. Excoecaria agallocha occurs throughout the 
forest but is relatively more abundant in areas which receive less freshwater. 
Ceriops candelleana occurs mostly in mixture with Heritiera and Excoecaria. 
However, pure patches of Ceriops are not uncommon. Sonneratia apetala occurs 
mostly in pure patches along water courses and is a colonizing tree species 
growing on newly accreted land. Xylocarpus mekongensis occurs in association 
with Bruguiera gymnorrhiza in damper places throughout the forest. Common 
understorey species, particularly in moist soils, are Cynometra ramiflora 
(Singra) and Amoora cucullata (Amur) . Nypa palms occur on the edge of water 
courses and are more abundant along small creeks than big rivers. The more 
saline western portion of the forest has a greater abundance of Ceriops and 
Phoenix paludoaa together with some Excoecaria. 

Heritiera fomes is economically the most important tree species in the 
Sundarbans. The timber is extensively used in construction, piling, and boat 
building. Heritiera poles are also being used as transmission poles, 
Heritiera fuelwood is of a high quality and is used in domestic cooking, 
small scale industries and brick manufacture. A hardboard mill in Khulna 
exclusively uses Heritiera fuelwood as raw material. 


fixcoecaria, on the other hand, produces a high quality pulpwood and is 
exclusively used as raw material in the only newsprint paper mill in the 
country. The species is also used extensively in the match industry. 

Ceriops candelleana is a high quality fuelwood and is also used as 
fencing and house posts. Xylocarpua mekongensis and Bruguiera gymnorrhiza are 
used in construction and furniture manufacture. Cynometra ramiflora, Amoora 
cucullata and Hibiscus tiliaceus all produce good quality fuelwood. Nypa 
leaves are used for thatching in rural areas. 


The first effort at generating quantitative information on the growing 
stock in the Sundarbans was made by Curtis during the formulation of his 
working plan in the lata 1920s and early 1930s. Data collected from a number 
of sample plots of H. femes, which were established between 1893 and 1911, 
were analyzed and the volume and diameter growths were assessed. Sample plots 
for measurement of growth of other species were also established during the 
preparation of the plan. Rough estimates of growth of Excoecaria agallocha, 
Sonneratia apetala and Avicennia officinalis were carried out through counting 
of annual growth rings. Volume tables were prepared for all commercially 
important species. The forest was divided into four quality classes. 

A detailed inventory of the Sundarbans was carried out in the late 1950s 
by the Forestal Forestry and the Engineering International Ltd. This inventory 
used aerial photographs and applied photogrammetric techniques combined with 
statistically controlled ground sampling. The forest was divided into four 
quality classes of 50+, 35-49, 20-35 and below 20 feet mean heights. Class 
3, corresponding to heights of 20-35 feet, was divided into two subclasses of 
25-35 and 20-25 feet mean heights. The 20-24 feet mean height sub-class was 
created because large areas in the Western and Southwestern part of the forest 
fall in this mean height category. The density of the forest was divided into 
four classes as foolows: 

A 90 - 100% crown closure 

B 75 - 90% 

C 50-75% 

D less than 50% " w 

Volume tables for commercially important species were constructed and 
a summary of the merchantable volume of different species was tabulated as 
shown in Table 1. 

ODA carried out an inventory of the Sundarbans in the early 1980s and 
published the results in 1985. Important information generated by the 
inventory is shown in Tables 2 and 3. 

Even though Ceriops is the third most frequently occurring species in 
the Sundarbans, the volume of the growing stock of the species has not been 
calculated during either of these inventories. 

According to the ODA study 65% of the forest had a crown closure of more 
than 70% while the Forestal inventory (1960) classified 78% of the forest as 
having a canopy closure of more than 75%. 


Table C.I: Summary of net merchantable volumes in the 195fc 
Utilization standard : To 4 Inch top D(o.b.) 




(ft 3 ) 



acre (f) | 


and over 

and over 

and over 



" 9" I 
o. a.o. 

Spcis t 





67 184 












449 229 












25 950 












15 515 












15 152 












19 389 












2 827 












595 249 







289 1 

Source : Forestal Inventory Report of the Sundarbans 

Table 6.2: Area of major forest types 

Forts t types 


miles 2 

Km 2 





































Source : ODA Inventory Report, 1985 


IfcbkfiJ: Merchantable volume of dffemt specks in the 198fe 



f 3 

m 3 

H. fames 

240 071 000 

6 796 000 

E. agallocha 
S. apetala 

63 320 000 
16 810 000 
55 763 000 

1 793 000 
476 000 
1 579 000 


375 964 000 

10 646 000 

* Others include A. officinalis, X. mekongensis, B. gymnorrhiza 
and X. granatum. 

Source : ODA Inventory Report, 1985 

It is difficult to correctly compare the growing stock in the forest 
from the two inventories, as trees of different sizes were considered as 
merchantable trees during the two inventories. However, using the ODA 
inventory data, Balmforth (1985) has calculated the decline in the volume of 
H. fomes and E. agallocha between the two inventories, which stands in the 
order of 40% for each species. 


The Sundarbans supports a very rich and diverse fauna which includes at 
least 43 species of mammals, 52 species of reptiles and amphibians (Hendrichs, 
1985; Mukherjee 1975) and over 186 species of birds (Salter, 1984), The 
Sundarbans is the last remaining natural habitat of the famous Royal Bengal 
Tiger, 350-450 of which are still reported to be present in the Bangladesh 
portion of the forest. 

The mammals of the Sundarbans include the Royal Bengal Tiger (Panthera 
tigris) , civets (Paradoxurus hemaphrodi tug and Viverra zibetha) , three species 
of wild cat (Fells spp.) , mongoose (Herpestes spp.), the Smooth otter (Lutra 
perspillata) , Spotted deer (Axis axis) , Barking deer (Muntiacus muntjak) , Wild 
boar (Sus scrofa) , Rhesus monkey (Macaca mulatta) , Gangetic dolphin 
(Platanista gangretica) , porcupine (Hystrix spp.) , squirrels, rats and bats. 

The reptiles and amphibians of the Sundarbans include the Estuarine 
crocodile (Crocodilus porusus) , the Bengal monitor and the Yellow monitor 
(Varanus bengalensis and V. sal va tor), various lizards and geckoes, the Rock 
python (Python molurus) , King cobra (Naja naja) , several species of marine 
and freshwater turtles including the Green turtle and Ridley turtle (Cheonia 
in/das and Lepidochelys olivaca, the River or Estuarine terrapin (flatagrur 
baska) , frogs and toads . 


The bird population of the Sundarbans includes at least eight species 
of kingfishers, including the large Brown-winged and Stork-billed kingfishers 
(Pelargropsis amauroptera and P. capensis) , the magnificent White-bellied Sea 
Eagle (Haliacetus leucograster) , which is quite common, the very rare Grey- 
headed Fishing Eagle (IcJityophagra ichtyaetus) and Pallas' Fishing Eagle (H. 
leucoryphus) . Brahminy kites (Haliastur Indus) are also a very common sight. 
Herons, egrets, adjutant storks, plovers, sandpipers and other waders are to 
be seen along mudflats and sandbanks. Doves, pigeons, sea gulls and terns are 
likewise numerous. Other birds include the Red jungle fowl (Gall us grallus) , 
several woodpeckers, the Red-ringed parakeet (Psittacula Icrameri) and mynas 
(Acridotheres spp.) . 

Compiled lists of all the mammals, reptiles, amphibians and most of the 
birds occurring in the Sundarbans can be found in Das and Siddiqi (1985) . 

The rivers, channels and creeks within the forest are rich in fish, 
crustaceans and molluscs, which are tapped regularly on a commercial basis and 
thus generate employment for a large number of people. 120 species of fish 
are reported to be commonly caught by commercial fishermen from the Sundarbans 
waters (Seidensticker and Hai, 1985) . 


The Sundarbans plays a very important role in the economy of the region. 
A large number of people in the neighboring districts are directly or 
indirectly dependent on the forest for their livelihood. In addition, the 
forest makes a substantial contribution to the economy of the country. The 
forest now comprises about 45% of the nation's productive forest and naturally 
constitutes the single largest source of wood in the country. The forest also 
supplies raw materials for a number of industries including a large newsprint 
paper mill and a hardboard mill. 

The forest has, so far, been managed primarily for timber, fuelwood 
poles and industrial wood (raw material) . However, the current management 
practice, which has been developed over a long time, allows entrepreneurs both 
large and small, to participate in income generative activities. Before a 
temporary moratorium was imposed on forest harvest in the Sundarbans in 1989, 
the average annual harvest of various products was as follows: 

Table 6.4: Annual production from the Sundarbans 



68 000 
182 900 

15 900 
106 450 

Source: Bangladesh Forest Department 

Timber (m3) 
Industrial wood (m3) 
Transmission Poles (no.) 
Fuelwood (tons) 
Nypa leaves (tons) 
Phoenix leaves (tons) 
Grasses (tons) 


It may be mentioned here that Nyp& palm leaves and grasses are used for 
thatching, Phoenix leaves are used for making house walls on which mud is 
plastered, and are used extensively by the poorer sections of the rural 
population. Fuel wood, Phoenix leaves and grasses are sold in small boat loads 
of 2*8 tons capacities for which permits are issued to individuals for each 
boat load. This allows a large number of people with small capital to involve 
themselves in income generating activities. The timber and Nypa coupes are 
also divided into a large number of small lots of different values thus 
allowing a large number of entrepreneurs to engage themselves. 

The Sundarbans is not just a conglomeration of trees like most other 
forests, it is an ecosystem which is also very rich in a number of other 
resources. The Sundarbans has a rich fishery resource and it provides vital 
breeding and nursery grounds for the fish, crustaceans and molluscs that make 
up the fishery in the head of the Bay of Bengal. It is estimated that some 
35% of the total marine fish catch in the Bay of Bengal is of species 
dependent upon this forest for some period of their life. 

The rivers, channels and creeks in the Sundarbans are also very rich in 
fishery resources. As has already been mentioned, a total of at least 120 
species of fish are reported to be regularly caught by commercial fishermen 
in the Sundarbans. The fishing within the forest takes place year round while 
fishing along the outer coast is seasonal and takes place between October and 
February when at least 10 000 fishermen establish temporary camps in the 
forest and fish along the coast. The fish are mostly dried and shipped to the 
market, A large number of fishermen are also involved in year round fishing 
in the Sundarbans. According to a recent report by AWB (1991) , about 150 000 
metric tons of fish are caught from the Sundarbans and adjacent waters each 

Mollusc shells are extensively collected from the forest and converted 
into lime by heating at high temperatures. According to the Forest Department 
Records, on an average, 3 150 metric tons of shells are collected annually. 

Flowers of mangrove trees in the Sundarbans produce high quality honey 
and on an average about 200 metric tons of honey and 50 tons of bee-wax are 
collected by individual honey gatherers annually. 

It is therefore obvious that the Sundarbans plays a very vital role in 
the economy of the region and generates a lot of diverse economic activities 
and provides a livelihood to a very large number of people in the region. 

An estimate of the number of people directly dependent on the forest for 
their livelihood is not known. However, an estimate in the late 1970s puts 
the number of people present within the forest on any day during the peak 
working season at 45 000. This figure has most probably increased since. The 
cumulative number of people entering the forest is much higher. In fact, 
according to Forest Department Statistics, about 300 000 people enter the 
forest each year but unofficial estimates suggest that the figure is likely 
to be much higher. According to a report by ESCAP (1987), the actual figure 
for direct employment by the Sundarbans is likely to be in the range of 500 
000 * 600 000 people for at least half of the year. 

These figures do not provide any .indication of the number of people 
involved in Sundarbans products related income generating activities outside. 


In activities such as wood processing in mills and factories, 
transportation, retail and wholesale trade, no assessment of the nuntoer of 
people involved has ever been made. However, the number is expected to be 
high. The aggregate of people who are dependent on the Sundarbans is high. 
This, together with the quantum of economic activity that the forest 
generates, makes it very important for the economy of the country. The 
revenue earned by the Forest Department from the forest is also very high 
compared to expenditure. Between the fiscal years 1984-85 and 1988-89 the 
average annual income from the forest was Bangladesh Taka 264.38 million (US$ 
7.0 million) and the average annual expenditure was Bangladesh Taka 21.6 
million (US$ 576 000) . 

The forest has a huge potential for tourism and provides excellent 
opportunities for outdoor recreation and is a paradise for nature lovers. The 
potentials for tourism, research and nature and conservation education have 
not been exploited. These activities have potentials for generating 
additional economic activities. 

Bangladesh is often battered by cyclones and sea storms which result in 
colossal damage to human lives and properties. The Sundarbans act as a buffer 
between the densely populated agricultural land and the sea, and protects the 
hinterland from major damages. The cyclones and sea storms have wrought 
destruction in the coastal regions time and again during the last thirty years 
or so. However, because of the presence of the Sundarbans between the sea and 
the habitation, no noticeable damage has ever taken place in the large area 
behind the forest. 


The present-day management practices in the Sundarbans were evolved 
locally over time and were not introduced from outside. The actual 
evolutionary process, which started at the beginning of the current century, 
was developed during the early and middle part of the century and is reflected 
in the management plans/schemes executed during this period. The evolution 
of the process makes interesting reading and has therefore been included in 
this document. 

The exploitation of the tree resources of the Sundarbans on a commercial 
basis dates back to the middle of the 16th century when a local king imposed 
a levy on the export of wood from the forest and used this as a regular source 
of income. Since then, local landlords realized tolls from the export of 
wood. In the early days of the British Indian Government, leases were granted 
to both locals and Europeans who were allowed to clear forests and convert 
them into agricultural land or human settlement. 

The value of the Sundarbans as a forest resource base was first realized 
in the 1860s, and in 1879 a Forest Division was created for the management of 
the Sundarbans. The first management prescription was developed after two 
reports on status of the tferitiera, the only tree species thought to be of 
commercial value and extensively exploited at that time, were made by A.L. 
Home (1872-73) and Dr. Schlich and Sir Richard Temple (1873-74). A minimum 
exploitable girth for Sundri was fixed in 1874 . Toll Collection Stations were 
also established on the recommendation of Dr. Schlich on the main routes of 
timber export. In the early days revenues were realized on the basis of the 
weight of the wood exported. 


The first management plan for the Sundarbans, which came into operation 
in 1893-94, was developed by R.L. Heinig who formulated management 
prescriptions for the exploitation of Heritiera trees in Bagerhut and Khulna 
forests where the forests were divided into 10 annual coupes for management 
purposes. The minimum exploitable diameter limit was refixed for Heritiera 
only which was thought to be the only commercially valuable species. However, 
regulated felling of trees continued in the rest of the forest* The main 
objective of the plan was to increase revenue and little attention was given 
to the silvicultural requirements or the conservation of Heritiera. This, as 
expected, resulted in substantial depletion of the forest. 

Luckily, the value of preservation of the forest took precedence over 
revenue generation at this stage and the follow-up actions which were taken 
laid the foundation for the evolution of a sustainable forest management 
regime, which today is still in practice, in a modified form. The adverse 
effect of over-exploitation was recognized before irreparable dapnage was 
caused to the forest and steps aimed at better regulation of felling of trees 
were taken. In the working scheme written by Lloyd for the period from 
1903-04 to 1908-09, the sizes of annual coupes were reduced to a quarter of 
the original size and the felling cycle was increased from 10 years to 40 
years. This resulted in the exploitation of a much lesser area annually and 
the gap between two fellings in any area was increased to 40 years. Simple 
silvicultural treatments were also prescribed. 

During this plan period all felling operations, including those of 
species other than Heritiera/ were confined to annual coupes and unregulated 
scattered felling was discontinued. Proper supervision of coupes was 
introduced and necessary measures were taken to control pilferage of wood. 

Traf ford's working plan which was written for the period 1912-13 to 
1931-32, divided the entire forest into two working circles: the Sundri 
(Heritiera) or Eastern Working Circle, covering relatively freshwater areas 
with good quality crop, and the Western Working Circle, covering more saline 
area where the crop was poorer. 

This plan fixed the exploitable girth for the other commercial species 
and introduced an intermediate type felling called the main thinning in order 
to relieve the congestion among middle aged crop. This prescription was 
designed to make improvement on felling and thinning on a 20-year cycle in 
more rapidly growing forests in the Eastern circle and 40 -year cycle in the 
other areas. Provisions were also made for subsidiary thinning among 
congested saplings. 

The working plan written by Curtis for the period 1931 to 1951 was a 
very comprehensive document which incorporated a number of modern management 
concepts, synthesized the experience from the past management practices and 
developed a management regime for perpetual supply of timber, fuel and 
thatching material. The prescriptions formulated by Curtis are still in 
practice and almost in the same form. Detailed enumeration, stock mapping, 
site classification, etc., were done and exploitable diameters for all 
commercial species were fixed for all site classes. 

Detailed rules and procedures for calculation of yield, laying out of 
coupes, marking and felling of trees including Nypa palms were made in the 
plan. These prescriptions are still followed. However, the plan was thought 
to be too intensive at the time and was simplified by S. Choudhury in 1937. 


Curtis also decentralized the forest management by creating six forest 
ranges. The forest was worked on a 20, 30 and 40-year cycle in good, moderate 
and poor quality forests. The yield was regulated by area even though 
provisions for a detailed yield calculation were made in the plan. Separate 
sets of rules were developed for the management of minor forest produces 
including Nypa palm leaves, honey, wax, grass and shells. 

A.M. Choudhury revised the plan for the then East Pakistan portion of 
the forest in the early 1960s after a detailed inventory was carried out and 
a lot of quantitative information was generated. Since a newsprint paper mill 
was then established in Khulna, and was designed to utilize Excoecaria (Qewa) 
as the pulping raw material, emphasis was laid upon the management of this 
species for the first time. A cutting cycle of 20 years was fixed for the 
entire forest and an annual allowable cut was calculated by using information 
generated from the inventory. Yield had, however, been regulated by area and 
while prescribing annual coupes the areas had been fixed so as to equalize the 
volume yield as far as possible in the case of the Heritiera (Sundri) forests. 
The rules laid down by Curtis were also fine tuned by Choudhury. 

Choudhury' s plan expired in 1980 and an inventory of the forest 
resources was carried out and a report was published in 1985. A revised plan 
is reported to be under preparation and meanwhile, the forest is being managed 
on prescriptions formulated on an annual basis. 

There has been a strong awareness in the recent years about the 
depletion of the growing stock of the two major species in the Sundarbans. 
Efforts are being made to restrict the annual Excoecaria harvest to 3.8 
million ft 3 . A complete ban was imposed in 1989 on the harvest of Heritiera. 
This order has since been modified and top-dying H, forties trees are being 
harvested from 1990. This annual harvest is located to one location in the 
forest and is undertaken by the forest department. So, the process of harvest 
of H. forties timber and fuelwood, as described in the preceding paragraphs, has 
temporarily been suspended. Harvest of other fuelwood, palm leaves and other 
minor produces continues as before. 


Administratively, the Sundarbans is under the control of one Divisional 
Forest Officer who is based in Khulna and is assisted in the discharge of his 
duties by an additional Divisional Forest Officer. The forest is divided into 
four ranges - Sarankhola, Chandpai, Khulna and Satkhira Ranges. The area and 
location of range headquarters are furnished below. 

Table 6.5: Ranges in the Sundarbus 

Hangs Headquarters Area (Ha) 

Sarankhola Sarankhola 130 998.0 

Chandpai Chandpai 100 021.0 

Khulna Na liana la 161 345.0 

Satkhira Burigoalini 184 992.0 


Bach range is under the management of a Range Officer who is a 
Professional Forester* The Range Officer is assisted by a Sub- Professional 
Assistant. There is a Game Warden who is in charge of the game sanctuaries. 
In addition to the above mentioned offices, a Forest Check Station has been 
established on the bank of each river and channel at the point where they 
enter the forest. There are 16 such stations in the Sundarbans. These 
stations primarily perform protective functions by checking the produce and 
ensuring that no unauthorized produce is brought out of the forest. These 
stations also regulate the flow of boats into the forest for collection of 
fuelwood which will be discussed later in the paper. There is no other 
permanent establishment inside the forest. Temporary coupe offices are 
constructed for supervision of activities near the coupes or the felling 
areas. These are so constructed that activities of 2 to 3 years coupes can 
be supervised from each office. A Sub- Professional Forest Ranger is normally 
in -charge of a coupe and is responsible for pre- fell ing marking of the coupes 
as well as the entire wood harvest operation. The Coupe Officers are assisted 
by a number of staff members in the discharge of their duties. These staff 
members include Deputy Rangers, Foresters and Forest Guards. An office is 
established in Dubla Island for revenue collection and supervision of fishing 
activities in the coastal waters in the winter months. This operation is 
headed by a Forest Ranger. 


The gradual fine tuning of management prescriptions initiated by Lloyd 
early this century has resulted in the development of a management regime 
which, if followed strictly, will ensure sustainable management of tree 
resources . 

The management practice currently followed was developed taking into 
consideration primarily the silvicultural requirements of the species and the 
need to maintain the crop in a steady and unchanged condition. The decision 
on the size of the .coupe or the degree or intensity of felling was based on 
a sample set of rules where the density, size and overall condition of the 
crop were the decisive factors in controlling the level of harvest. These 
easy-to-follow rules ensure that the felling practice does not cause any 
depletion of the forest. Even though two very detailed inventories have been 
carried out in the Sundarbans, very little information is as yet available on 
the rate of growth and yield from the forest. Such information is a basic 
pre-requisite for formulating sustainable management regimes based on 
quantitative information. 

The management practices in the Sundarbans involve a one-time harvest 
and improvement felling of tree resources in any portion of the productive 
forest, once in each 20-year cycle. This operation normally lasts for about 
a year to a year and a half and after that the area is left undisturbed till 
it is time for another harvest in the next cycle. This allows the forest to 
grow undisturbed without any outside interference and causes minimum possible 
disturbance to the wildlife. Since an annual coupe is scattered over the 
entire forest, the size of the area under exploitation in one center at any 
one time is also very small. The rules which are followed in timber harvest 
operation are described in Box 1: 



* * . . '' ' *!i ' v ' '' *'''&' .'. -I. .Si/ '* :'/. /jH : i^. . ! 

"""" ' ' " " $i 

canopy, except in places where regenerat ion h* 


ultitnately become 

Removal of diseased and large spreading, misshaped 
even if this creates gaps. 

The improvement felling consist* of 

felling of some trees in yotmger crop for 
of congestion in the crop and 



Box 1 : General timber harvesting rules 

removal of diseased, dead and dying trees for 
of the hygienic condition of t|i9 


I #:- ' - 

These rules apply to both Heritiera and Excoecaria except that in the 
case of Excoecaria no improvement felling is undertaken. 

In the case of Ceriops, the felling in a given year is confined to an 

annual coupe area and the felling process involves three steps: 

(i) selective felling of Ceriqps poles 

(ii) removal of other exploitable Ceriqps trees 

(iii) thinning of congested crop. 

These rules are simple and easy to follow even though they require some 
subjective judgement in the selection of trees. This regime does not follow 
the basic principle of sustainable management where removal is equal to the 
increment of the forest during the period between the previous and the current 
fellings. However, what it does do is to disturb the forest to the minimum 
and to allow the removal of only those trees which have reached maturity and 
whose removal results in the occupation of the vacated space by new 
recruitments. Such fellings, together with improvement fellings, create 
better conditions for the enhanced growth of the remaining trees. 

In the absence of detailed information on growth and yield, the 
management practice is ideal for maintaining the forest in its original 
condition. It also ensures a sustainable production when coupes are laid in 
such a way that annual coupes provide more or less equal output. This is 
quite possible because of the detailed information which is now available on 
the sizes and distribution of trees of different species in different areas 
within the forest. 

Different methods are employed for the disposal of different produces. 
Based on the management prescriptions, annual coupes are laid out at different 
locations in different ranges in the forest. 


Normally coupes are laid out in one or two locations in each range. The 
area at each site is divided into small lots of 10-20 acres (4-8 ha) by 
clearing vegetation in lines along a North-South and Bast -West direction. The 
trees are then measured and marked following the prescription described in 
Section 6. A list showing the species, diameter and height is made for each 
lot. Any deformities in the trees are also mentioned. A consolidated list 
of all lots is then published and a date for the sale of the lots fixed and 
announced both in local and national newspapers as well as the government 
gazette. Lots are sold in open auction to the highest bidder. The purchaser 
then enters into an agreement which details the payment schedule and other 
terms and conditions. The purchasers are allowed about 9 months to complete 
the harvest. 

After the timber extraction has been completed, the Coupe Officer makes 
an assessment of fuelwood that will be available from thinning and improvement 
felling in each lot and informs the Range Officer. Normally, the Coupe 
Officer sends a consolidated statement of how much fuelwood is available from 
the officer's jurisdiction each fortnight. 

Fuelwood sales follow an altogether different mechanism. Heritiera 
fuelwood is sold in small lots of 1,5 to 6 metric tons by boat loads to 
individual wood collectors. Individual woodcutters commonly known as 
'Jbawalis' register their boats at any Forest Check Station along the boundary 
of the forest. The registration is a two step process. The first step 
involves the registration of the Boat and procurement of a Boat Licensing 
Certificate (BLC) . This certificate provides the name and address of the 
owner together with the dimensions and load carrying capacity. The load 
carrying capacity of a boat is calculated using the following formula. 

Load carrying capacity* = L x B x D x 0.356 

where L Total length of the boat 

B - Breadth at the widest point 

D * Depth at the deepest point 

0.356 * a constant 

* Load carrying capacity in maunds 
(1 maund 37 kg.) 

Once the BLC has been procured, the woodcutters register their boats 
with the station for fuelwood collection and are given serial numbers. They 
then wait their turn for permission to proceed to a coupe for fuelwood 

Once the Range Officers receive the fortnightly return of fuelwood 
available in the coupes, they allocate them to different Forest Stations where 
the station officer issues permission to waiting woodcutters serially from the 
waiting list. 

Woodcutters then proceed to coupes, pay the royalty and are assigned 
lots for collection of wood. Separate permits are issued for dry lops and 
tops left behind after the timber harvest operation and green fuelwood from 
the thinning and improvement felling. 


Separate coupes are laid out for the extraction of Ceriopa poles and 
fuel wood. Coupe Officers issue permits on a first -come -first -served basis, 
allocate areas for harvests and supervise the operations. 

While extracting trees, care should be taken to ensure that at least one 
Ceriops shoot, preferably more, should be left on each root or bunch. Except 
where absolutely unavoidable, woodcutters should not be allowed to cut young 
withes merely to facilitate felling, and not for actual utilization. Solitary 
stems of under 2 inches in diameter at 3 feet from the base, should be left 
to produce either a larger pole or a larger root and more shoots. 

The Newsprint Mill carries out its own operation for the extraction of 
the Excoecaria pulp wood. The Forest Department allocates the area and 
normally supervises the operation. The Mill has developed an elaborate set-up 
for the operation as described in Box 2 below. 

U) Felling plans covering the geographical and chronologic*! 
distribution of the Newsprint cutting operation is prepared 
by the Forestry Section of Khulna Newsprlrtt Mill, tl^ffe 
felling plans cover periods of 5 years, 

(ii) where it is considered aecessai^r, in the eatinwition of the 
Divisional Forest Officer, sufficient sound, healthy and 
well shaped Excoecaria trees above 4,6" dtoh will be marked 
with paint for retention as seed bearers. 

Uii) Felling of a group of trees in 
enlarging an exiifltifiigf gap id to 

regeneration is established. This rule does not place an 
embargo on felling of individual large and spreading trees 
or any diseased tree which will eventually cause a gup when 

<iv) All fixtfoeearia trees of 4*6" Sbh and above and not covered 
by the restrictions ta rule (ill or UU) are 

Box 2 : Umber harvesting rules used by the Newsprint Mill 

Excoecaria is also used in the matchwood industry. A separate set of 
felling and marking rules applies in this case. These are described in Box 3. 

Inspite of the above practices, there has been a major depletion in the 
merchantable growing stock between 1960 and 1985 when two inventory reports 
were published. Even though the causes of the depletion are not documented 
or identified, the system of management cannot be faulted for this depletion, 
since, if followed strictly, the system offers no scope for depletion. 

The depletion of Heritiera may have resulted from over -exploitation 
particularly during improvement felling which accounts for up to 85% of all 
the removals (Balmforth, 1985). This operation is carried out with very 
little supervision, if any, and is left to low level forest staff who rosy not 
have the knowledge or concept of how this operation should be carried out. 


in' the" 

raanage*nt plan, Ea<A <^mpe is divided into s^tioiis of 
appr<mitwtely 40 acres each by making North-South and g**t- 
West lines 25 chains apart . With the aid of these lines, 'the 
will 3be w^wwS 9** a f $&le 4 inphe* tp 1 wile. 
of the small creeks will be shown on this map. 

(ii) All unsound and badly shaped or otherwise defective trees are 
marked, provided their removal does not create a permanent 
gap, Diseased trees are retr^ved under any circumstances. 

(iii) All trees 6" diameter and above at breast height are marked 
for felling. 

(iv) Marking of a group of trees in one place and the enlargement 
of an existing blank area is to be avoided, except where 
regeneration has been established. 

(v) Felling of one am of a forked tree is to be avoided as the 
remaining arm generally becomes unsound. 

(vi) All trees should be hammer-marked at a height of 4* - 6" and 
at the base. The base mark should be as low down as possible 
to avoid waste. 

(vii) iti*t of trees marked should be prepared according to 1" 
diameter classes. 

(viii) Utilization will be to 4" top diameter and all attempts 
should be made to avoid any waste. 

Box 3 : Umber harvesting rules used by the Match Factories 

Frequent cyclone and death of trees because of a die back disease (the 
reason of which is still unknown - refer to BARC, 1990) , account for some 
depletion of Heritiera. Pilferage of Heritiera is also becoming a matter of 
serious concern. 

The cause of depletion of Excoecaria stock is, however, clearer. Over- 
exploitation of the growing stock has continued from the time of the 
establishment of a newsprint paper mill in the late 1950s. This public sector 
undertaking had been enjoying the privilege of harvesting the entire pulping 
raw material requirement from the forest without due consideration to the 
sustainability of such actions. 

The mill carries out its own wood harvest operation with very little, 
if any, on the ground supervision from the Forest Department. The Forest 
Department allocates the felling area and fixes the annual quota. The ODA 
inventory has recommended an annual harvest of 2.5 million cubic feet while 
records show that 8.9 million cubic feet have been harvested in 1984-85. Even 
though the harvest had fallen to 4.7 million cubic feet in 1989-90, it is far 
above the ODA recommended quantity. 

The Forest Department has been trying to limit current felling of 
Excoecaria to 3.8 million cubic feet, but this is also well in excess of the 
quantity recommended* However, despite this over-exploitation, regeneration, 
which has occurred in these local areas, remains satisfactory. 


Figure 6.2: Bundling of Excoecaria wood 
Photo by M.L. Wilkie 


t-. .,... ..JlJP'wto^frs-iJ. 

y-^iaft ^L-SiBfc"'^ ' i ' 

Figure 6 J: Excoecarie wood ready to be transported to the Newsprint MiB 

Photo by M.L. Wilkie 


Permits are issued from check stations on a first-come-first-served 
basis for the extraction of Amoora, Cynometra and Hibiscus fuel wood. 

Jfypa leaves, or Golpatta as it is commonly known! is an extensively used 
thatching material in Southeastern Bangladesh. The permissions to harvest the 
leaves are sold in open auction and the purchasers divide these permissions 
up and sell them to actual collectors by small boat loads. For the purpose 
of management, each year's working area is divided into a number of coupes 
(usually seven) and each is sold separately. The Nypa extraction is now 
regulated completely by the Forest Department through Coupe Officers who issue 
permits for collection and supervise the operation. The extraction is carried 
out in winter months. 

The rules outlined below in Box 4, which were originally formulated by 
Curtis are still followed. 



$;.fc;rte* should be exploited twwe than once a year* 

New fr<m<Jr or so-called "central leaves* should not be cut; 
also purchasers must not be allowed to cut leaves which they 
do not intend to utilize, but to leave on the ground to rot. 
Young plants with only one utiliz*ble leaf should not be cut. 

The main work of Coupe Of fleers of Golpatto coupes will be to 
see that rules (1) and {2} are obeyed, and that no Golpatta 
in the interior of the forests is left unworked before the 
coupe moves on. Each purcha&er should be allotted a small 
khal, or a part of a large khal to work in, and should not be 
given a fresh area until the area already allotted is 
finished. Areas near the sea- face should be worked during 
the eali season. 

1^ the forests, Qouj>e Officers will 

prepare stock maps on a 4 -inch scale of the Golpatta in each 
^^<i^fc^fiitt,/ ' i -^:;^;^'''feoa : ia toe aeafc to tfee eivistctoel 
Of f ice for record, and one copy kept on the coupe to 


Box 4: Harratiiif rules for Nypa (GdpttU) froo* 

Grasses and Hantal (Phoenix paludosa) leaves are also sold by permits 
to individual collectors and are regulated by Station Officers at the entrance 
of the forest. 

Some of the mangrove trees produce excellent quality honey and permits 
for honey and wax collection are issued to individuals between 1 April and 15 

Hie major non- fores try products of the Sundarbans are different 
varieties of fish, prawn and molluscs. The harvest of these items are 
controlled from Forest Stations. Permits are issued for a specified period 
of time and collection or harvesting is carried out by permit holders. 


Figure 6.4: A boat loaded with Nypa leaves 
Photo by M.L. Wiltie 

Figure 6.5: A fishing boat and crew 

Photo by M.L. Wilkie 


A fishing camp is operated on Dubla Island, just in- shore from the Bay 
of Bengal, from November to February each year. Fishermen establish temporary 
camps on the island under the supervision of the Forest Department, catch fish 
from the Bay of Bengal and mostly dry them for export. Some fresh fish are 
also marketed. The Forest Department collects tolls from the fishermen* 

Except for Nypa leaves, no management practice is in place for the 
sustainable management of other minor and non-forestry products. The number 
of permits to be issued for any one item is also not regulated. But it is 
interesting to note that there is no trend of decline in annual catch or 


The management of wildlife is confined to their protection against 
poaching. Hunting is strictly prohibited in the forest except in cases where 
it becomes necessary to kill a man-eater (tiger) . 

Three wildlife sanctuaries have been established in the Sundarbans. 
Additional protection is provided to wildlife in these forests. No census of 
wildlife in the Sundarbans has ever been carried out. However, inspite of the 
very low level of management, the fauna of the Sundarbans has remained pretty 
much intact during the current century. 


The Sundarbans is not just a conglomerate of trees stretching over a 
large area. It is a unique ecosystem, the components of which are very 
diverse. It is the habitat of the largest remaining population of the famous 
Royal Bengal Tiger (Panthera tigris) . In addition, it is the home of a 
diverse fauna which includes other endangered species such as the Lesser 
Adjutant Stork (Leptoptilus javanicus) ,the Estuarine crocodile (Crocodilus 
poroaua) , the Rock Python (Python molurus) and the Estuarine terrapin (Batagur 
baska) . 

The Sundarbans plays a very important role in the economy of the region 
and according to an ESCAP (1987) estimate, about half a million people may be 
entering the Sundarbans each year for income generating activities. There is 
also a large number of people in primary and secondary processing, retail and 
wholesale trade, who are dependent on the Sundarbans for their livelihood. 
It is, therefore, very important that the conservation of this forest in its 
original condition is ensured at all costs. 

In the future management of the Sundarbans, the conservation of the 
forest should be considered as the primary factor in formulating management 
regimes. Production from the forest should be of secondary consideration and 
limited only to the quantum which will not cause any depletion of the 
resource. The Sundarbans is a unique gift of nature, not only to Bangladesh, 
but to the whole world, and it is our solemn duty to conserve and maintain it 


Figure 6.6: Spotted deer coming down to drink at a man made water hole 

Photo by M.L. Wilkie 

Figure 6.7: The Sundarbans - a mangrove forest with great biodiversity 

Photo by M.L. Wilkie 


AWB, Management of Sundarbans Reserved Forest in Bangladesh Forestry 
1991 III Project* Environment Component, Asian Wetlands Bureau. 

Balmforth, E.G. Preparation of Interim Felling Prescriptions for the 
1985 Sundarbans Forest. FAO, Dhaka. 

Chaffey, D.R., F.R. Miller, J.H. Sandom. A forest Inventory of the 
1985 Sundarbans, Bangladesh. ODA, Surrey, England. 

Chaudhury, A.M. Working Plan of Sundarban Forest Division for the Period 
1968 from 1960 - 61 to 1979 - 80. East Pakistan Government Press. 

Chaudhury, R.A. and M.Z. Hussain. Forest Management Practices in 
1989 Bangladesh - Traditional Practices and Alternative Approaches. 
In PROC. National Seminar on Forest Resources Management in 
Bangladesh - Issues Problems and Prospect (S.H. Rahman, M.Z. 
Hussain, S.I. Ali and S. Hugs Eds). 

Curtis, S.J. Working Plan for the Forests of Sundarbans Division. Bengal 
1933 Government Press, Calcutta. 

Das, S. and N. A. Siddiqi. The mangroves and mangrove Forest of 
1985 Bangladesh. Mangrove Silviculture Division Bulletin No. 2. FRI 

ESCAP Coastal Environmantal Management Plan for Bangladesh. Volume 2 

1987 Final Report. ESCAP, Bangkok, Thailand. 

Forestal Forestry and Engineering International Limited. Inventory of the 
1960 Sundarbans Forests. Vancuver, Canada 

Hendrichs, H. The status of the tiger (Panthera tigris) (Linne) in the 
1975 Sundarbans mangrove forest (Bay of Bengal) . Saugetier- 

kundliche Mitteilungen. 23: 161-199. (Reported in Das and 

Siddiqi, 1985) 

Hussain, M.Z. Development of Forest Resources in Bangladesh. Forest 
1991 Department. Dhaka 

Hussain, M.Z. Sustainable Management of Forest Resources in Bangladesh. 
1991 MARC, Dhaka. 

Karim, A. Environmental Factors and the Distribution of Mangroves of 

1988 Sundarbans With Special Reference to Heritiera fames. Doctoral 
dissertation, University of Calcutta. 

Mukherjee, A.K. The Sundarbans of India and its biota. J. Bombay Nat. 
1975 Hist. Soc. 72 : 1-20. (Reported in Das and Siddiqi, 1985) 

Seidensticker J, and Hai A. The Sundarbans wildlife management plan. 
1985 IUCM, Gland, Switzerland, 120 pp. (Reported in ESCAP, 1987) 

Zaman, M.N. Sundarbans at a glance (in Bengali) . 

1989 Forest Department. Khulna. 






A.I.I Aerial photography 261 

A. l.l.l General characteristics of aerial photography . 261 

A. 1.1. 2 Oblique and panoramic photography 262 

A. 1.1. 3 Small format photography 263 

A. 1.2 Airborne video systems 263 

A, 1.3 Satellite imagery 264 

A. 1.3.1 The Landsat system 264 

A. 1.3. 2 The SPOT system 265 

A. 1.4 Radar system 267 


A. 2.1 Aerial photographs 268 

A. 2. 1.1 Pictorial characteristics 268 

A. 2. 1.2 Photo- interpretation keys 269 

A. 2. 1.3 Photo- interpretation aids . . 269 

A. 2. 2 Airborne video systems 270 

A. 2. 3 Satellite imagery 270 

A. 2. 3.1 Analog interpretation of satellite images ... 270 

A. 2. 3. 2 Computer-aided classification 273 

A. 2. 4 Radar imagery 276 



A. 3.1 Mapping procedures 277 

A. 3. l.i Maps from aerial photography 277 

A. 3. 1.2 Maps from satellite imagery 279 

A. 3. 1.3 Map accuracy 280 

A. 3, 2 Mosaic compilation 285 




A. 4.1 Stratified sampling 286 

A. 4. 2 Multi-phase sampling 286 

A. 4. 2.1 Double-phase sampling 286 

A. 4. 2. 2 Three-phase sampling 289 

A. 4. 3 Cluster sampling 292 



A. 5.1 Latin America 293 

A. 5. 2 Africa 294 

A. 5. 3 Asia 295 




A.l.l Classification of scales used in forestry 261 

A. 1.2 Example of lateral and areal format coverage 262 

A. 1.3 Spectral sensitivity and resolution of Landsat sensors .... 266 

A. 1.4 Characteristics of SPOT sensors 266 

A. 1.5 Comparative table of different sensors in terms of wavelengths 

and resolution 267 

A. 3.1 Examples of simple transfer instruments 278 

A. 3. 2 An example of an error matrix 283 

A. 3. 3 An example of an error matrix with area values instead of point 284 


A. 5.1 Volume regressions for Keora (Sonneratia apetala) 295 

A. 5. 2 Volume regressions for Gewa (Excoecaria agallocha) 296 

A. 5. 3 Volume regressions for Sundri (Heritiera fomes) 297 


A. 2.1 Enhancement techniques for analog photo- interpretation .... 271 
A. 2. 2 Enhancement techniques for computer aided interpretation . . . 275 



Aerial photography has applications in many forestry related activities 
including forest type mapping, tree species identification, drainage and 
land-use classification, forest inventory and forest management plans, road 
site locations, fire and disease damage assessment, erosion studies etc. 

Depending on the information needed and the ground resolution required 
- which is directly related to the classification precision - characteristics 
of aerial photography should be specified. These characteristics concern the 
photographic system, the type of film, the scale and the atmospheric 

A. 1.1.1 General characteristics of aerial photography 

The choice of a photo scale depends on the area to be covered, 
the required ground resolution and the coverage cost. For forest 
inventory operations, different scales also depend on the purpose of 
the survey. Table A. 1.1 below gives a simplified classification of 
scales used in forestry. 

Table A.U: Classification of scales used in forestry 


Vary large scale 
Large ecale 
Medium scale 
Small scale 

> 1:5 000 

1:5 000 - 1:10 000 
1:10 000 - 1:30 000 
1:30 000 - 1:60 000 

Very email ecale 1:60 OOP and smaller 


Tree enumeration - damage assessment 
Stand measurement for inventory 
Forest type mapping - management planning 
Topomaps - Reconnaissance surveys 
National forest inventory 

Source: Remeijin (1981). 

In the extensive tropical forests, including mangroves, small 
scales, down to 1:60 000, might be used for broad type mapping. As 
aerial photography of about 1:50 000, designed for topographic maps, is 
generally available, it can also be used for broad forest type 
delineation and for planning forest inventories. 


Together with the photo scale, the format is an important element 
to consider during a photographic coverage. The most common format 
available is 23x23 cm (9x9 in) but photos of 18x18 cm are also used 
occasionally. Small format photography, obtained from 70 mm and 35 mm 
cameras (57x57 mm and 24x36 mm respectively), are also increasingly 
used in forestry applications. These particular systems are presented 
in a later section. 

In relatively flat areas such as mangroves, large formats are to 
be preferred since no important relief displacement occurs on the photo 
image. In combination with the format, one should also consider the 
camera focal length to be used. 


With a 23x23 cm format for example, a camera focal length of 152 
mm will provide a wide angle photograph usually used for mapping over 
flat terrain. Wide angle photography, however, causes difficulties in 
seeing the details on the edges of the prints. Table A. 2. 2 below gives 
examples of lateral and areal format coverage at several scales. 

TaUe A ,1.2: Example of lateral and aretl format coverage 



Lateral coverage () 

Araal coverage (Ha 

3-5 m (24x36 ran) 

1 2 500 



70 m (57x57 mn) 

1 2 500 
1 5 000 



9x9 in. (23x23 cm) 

1 12 000 
1 30 000 
1 000 
1 120 000 



Source: SAF Forestry handbook (2nd edition). 

To secure complete photo coverage, necessary for mapping and for 
stereoscopic observation, end-lap (overlap between successive photos 
within a flight line) and side-lap (overlap between flight lines) must 
be about 605% and 3015% respectively. It should also be recalled 
that crab and drift, due to bad navigation, may cause insufficient end- 
and side- lap, which results in gaps between flight lines. 

Type of fila 

The most commonly used types of films are panchromatic, infrared 
black and white, colour and colour infrared. Their constitution and 
spectral sensitivity range are described in various text books and 
documents. It should be noted however that species identification and 
detection of diseased trees are much easier, on colour and infrared 
colour films photography than on panchromatic emulsions, in spite of 
their higher cost. 

Various other photo characteristics must be carefully considered 
when a photo survey is being planned. They concern film processing, 
paper type, timing, storage, etc. 

A. 1.1. 2 Oblique and panoramic photography 

The interest of oblique photographs resides in their use for 
reconnaissance survey and illustrative purposes. Compared to vertical 
photographs, they cover a much larger area. However, they present the 
disadvantage of not having the geometric accuracy required for mapping 
and measurements. 

Panoramic photographs are also used in reconnaissance surveys. 
They are characterized by a wide angle field of view and high 
resolution. Like oblique photographs, they are not suitable for 
mapping because of the geometrical distortion of the image. 

A. 1.1. 3 Small format photography 

The most commonly applied systems utilize a motor driven 35 mm or 
70 mm camera which can be mounted on a light aircraft or a helicopter. 
The system can operate with different film/lens combinations. 

The quality of contemporary 35 or 70 mm cameras and films is such 
that prints enlarged from negatives can provide acceptable detail. The 
low cost of such systems and the simplicity of the equipment has drawn 
the attention of many researchers and inventory groups. Small format 
photography has been used in many applications in forestry such as 
detection and evaluation of forest disease, monitoring land management 
activities and in forest and range inventories - especially in areas of 
difficult access. 

Gains from small format photography systems might be expected in: 

Rapid data acquisition, i.e. a short period of time is required 
to assure up-to-date photo coverage; 

Timing, i.e. the simplicity of the system allows for rapid 
preparation of flight planning and execution and requires 
relatively simple film processing arrangements; 

Small format photography is an excellent sampling tool. It 
requires relatively inexpensive photographic and photo- 
interpretation equipment, and provides fine resolution at low 

However, the system presents a major disadvantage of having a 
limited areal coverage per frame and it is subsequently not suitable 
for complete coverage of large areas unless the photo scale is small. 
Moreover, for precise flying height determination special devices 
(laser or radar altimeter) are required. 


Recent attempts have shown the advantages of video systems in vegetation 
assessments. The system is usually loaded on a light aircraft in the same 
manner as small format photography using a camera mount. The viewing system 
is mostly composed of a camera, a monitor and a video-cassette recorder and 
can operate in an autonomous manner if a battery pack is incorporated in the 

Black and white, colour or even colour infra red imagery can be 
produced, allowing for a large spectrum of applications. Due to fast 
developments in the techniques of recording and the improvement on the 
resolution, this relatively recent technique is gaining much interest among 
researchers concerned with vegetation studies. When applied to forest 
mapping, the system has been found to facilitate the distinction between 
forest types based on colour combined with crown contour and size (Vlcek, 
1983) . 


Advantages of the video system reside in: 

The relatively low cost of imagery when compared with photographic 
systems such as a 35 mm camera setting; 

The system may be quickly set up for operation and for a quick 
evaluation after coverage; 

It also presents the ability of yielding a real time imagery during the 
flight which allows the operator to interactively adjust the camera 
and the tape recorder settings; 

Due to the digital format of video data, the system can potentially be 
used in computer image processing; 

The incorporation of an audio track allows for oral comments to 
complement vegetation description during the flight. 

Considering the image sharpness and definition and the quality of the 
product/ major actual disadvantages of video imagery is its poor resolution 
and difficulties in establishing the scale of the images. In spite of these 
drawbacks, such a system can be useful in reconnaissance surveys and broad 
forest type classifications. 


The application of satellite imagery to vegetation studies - 
particularly in mangrove areas - has mainly been, and still is, using Landsat 
data and more recently SPOT imagery. The latter is increasingly being 
introduced in survey operations. Its major advantage resides in a higher 
ground resolution. 

In the following sections, some basic technical aspects of major sensors 
on board satellites are presented. The indications outlined are those mostly 
related to vegetation mapping. More details on system configuration, orbit 
characteristics and image processing can be found in various textbooks and 
documents (ASP, 1983; Baltaxe, 1980 and many others). 

A. 1.3.1 The Landsat system 

The characteristics of the Landsat satellite system has undergone 
continuous changes due to progressive modifications of different 
elements of the system and particularly the sensors on board since the 
launching of the first satellite of the series in 1972. 

The Haiti Spectral Scanner (MSS) 

The MSS records the solar radiation reflected by the earth's 
surface in four spectral bands. These bands are called band 4, green 
{0.5 - 0.6/jm), band 5, red (0.6 - 0.7/im), band 6 and band 7, reflected 
infra-red (0.7 - 0.8fzm and 0.8 - l.ljxm) respectively. 

On board Landsat 3, Launched in March 1978, major changes were 
introduced, namely the addition of a thermal channel which operates in 
the range 10.4 to 12.6 /im. 


Shortly after the launch of this satellite, the fifth channel was 
de-activated due to problems encountered with the quality of the data. 
The remaining four spectral channels continued to function normally 
until the retirement of the satellite in September 1983. In Landsat 4 
and 5, the spectral wave lengths covered by MSS data are the same as 
those in Landsat 1,2 and 3, and the ground resolution has also been 
kept at about 80 m. 

Tht Rtturn Btaa Vidicon (RBV) 

The images from the RBV have greater inherent cartographic 
fidelity than the Landsat MSS and contains a reseau grid which is 
inscribed on the photo conductive surface of the RBV tube. This reseau 
pattern facilitates the geometric correction of the imagery during the 
image recording process. Due to problems with the tape recorder on 
Landsat 1 and 2, the RBV was shut off and little RBV data are 

On Landsat 3, only two cameras were mounted compared to three on 
Landsat 1 and 2. The new system provides broad band imagery (0.505 - 
0.750/im, from green to near IR) with an improvement in ground 
resolution compared to the previous RBV. This arrangement makes two 
successive scene pairs (4 images) coincide with one MSS scene. These 
cameras have a 40 m spatial resolution or about twice that of the MSS 

Th* ThtMtic Mmpptr (TM) 

The Thematic Mapper is a seven-channel scanner which is designed 
to improve vegetation analysis capabilities. It operates with a 
spatial resolution of approximately 30 m which corresponds to about 2.6 
times the resolution of the MSS. Due to the great number of spectral 
bands, the TM is characterized by a greater radiometric sensitivity, 
Tablft A. 1.3 gives a summary of sensor characteristics in Landsat 
satellites launched so far. 

A. 1.3. 2 The SPOT system 

This system operates in the spectral bands (the visible and the 
near IR portion of the spectrum) with a ground resolution in the order 
of 20 m in the MSS mode and 10 m in the panchromatic mode. Tabl* X.I. 4 
indicates spectral sensitivity and ground resolution of the system. 

During the flight, a steerable mirror views the ground surface 
with a swath width of 60 to 80 km. With such a steerable device, two 
different viewing systems - vertical and lateral - are possible. The 
lateral viewing confers to the system a distinctive advantage of being 
able to increase the revisit frequency of a given scene on the earth's 
surface. This is significant as it provides the possibility of 
stereoscopic viewing of the image. 

Table A. 1.3 provides a summary of major sensors used on board 
satellites for vegetation studies. 


TtWeA.U: Spectral Maturity and resolution of Lands* 





Landsat no. 







0.5 - 0.6 





0.5 - 0.7 





0.7 - 0.8 





0.8 - 1.1 




10.4 - 12.6 





.475 - .575 




.580 - .680 




.690 - .830 


1*,2*,3 1 



.45 - .52 




.52 - .60 




.63 - .69 




.76 - .90 




1.55 - 1.75 




2.08 - 2.35 




0.40 -12.50 



*: non operational 

Table A. 1.4: Characteristics of Spot sensors 


Spectral bands 

Ground sampling 

Number of pixels 
per line 

Ground swath 


.50 - . 

.61 - .68/xm 

,79 - .89/im 


Source: Borel,1985 


Table A.1.5: Comparative table of different 







.38-. 42 



.42-. 45 



.45-. 50 

TM1 .45-. 52 

.50-. 55 


.55-. 60 

TM2 .52-. 60 

MSS4 .5-. 6 

SI .50-. 59 

.60-. 65 

.65-. 69 

TM3 .63-. 69 

MSS5 .6-. 7 

S2 .61-. 69 

.70-. 79 

MSS6 .7-. 8 


.80-. 89 

TM4 .76-. 90 

MCC7 ft - 1 1 

S3 .79-. 90 


PlQO / . JL J. 

TM5 1.55-1.75 

TM6 2.06-2.35 

TM7 10.4-12.5 







Source: Borel,1985. 


Side Looking Airborne Radar (SLAR) is the most common sensor operating 
in the microwave portion of the electromagnetic spectrum, which has been used 
for vegetation studies. SLAR has been more widely used in tropical forests 
because of its great ability of penetrating clouds and rains. This property 
makes radar system the best time independent sensor available (Morain, 1976) . 

Normally, resolution with radar is poorer than with sensors operating 
in the optical and IR regions. With the inclusion of radar in the pay load 
of space platforms, this system will play an increasingly important role in 
natural resources studies. The ground resolution of SLAR depends on the pulse 
rate and the antenna beam width. 




A. 2. 1.1 Pictorial characteristics 

The ability to identify terrestrial objects depends on various 
image characteristics but also on the objects to be interpreted. 
Stratification criteria used in mangrove studies are usually based upon 
the physiognomic appearance of the vegetation and on physiographic 
features such as landforms. Qualitative and quantitative pictorial 
elements which are used in the photo- interpretation process are 
extensively discussed in most photo- interpret at ion manuals. In short 
they include: 

Tons and colour 

The difference in tone or colour between an object and its 
environment on an image help identify the object. The contrast in grey 
tones and colours are determined by the intensity of the radiation 
reflected by the objects, the atmospheric conditions/ and the 
processing of the film. For forest classification, tones may depend on 
the sun angle during the time of exposure, the age of the trees, the 
moisture conditions and the stand density. That is why photo- 
interpretation should not be based on this single pictorial element 
alone. Moreover, differences of colour hues on colour prints should be 
carefully interpreted as perceived differences might not correspond to 
different vegetation types or trees. 

Siza and ahapa 

The size of an object on an image depends upon the photo scale, 
the object's size on the ground and the resolving power of the imaging 
system used. It is often the combination of size and shape which is 
responsible for the recognition. For forestry applications, the size 
and shape of tree crowns are useful in tree species identification. 

Taxtura and pattern 

The texture can be smooth, fine, coarse or rough and is very much 
dependent on the scale of the image. The pattern is the spatial 
arrangement of objects and their repetition on the image. The pattern 
can be natural or man made. Natural pattern usually results from local 
topography, drainage pattern etc., while a man made pattern is found in 
plantations, shifting cultivations, cutting regimes etc. 

Location and association 

Location and association are not object characteristics, but are 
related to the surroundings of the objects. The knowledge of the 
ecological conditions and environment in which an object is located, 
can be helpful in identifying the object. Also, a given object can be 
recognized simply from another neighbouring known object. 


The Nipa palm for example can be easily recognized because it is 
associated with mangrove forests and does not occur on dry sites 
(Boonchana, 1983) . 

Quantitative tltMntf 

Criteria of this type include all measurements which can be 
performed on aerial photographs. Recorded information is expressed 
quantitatively and is used to emphasize qualitative photographic 
information and improve classification accuracy. Such measurements are 
usually crown density, crown diameter and tree height. The accuracy of 
photo measurements depends on the photo scale, the size of the object 
and the instruments used. 

A. 2. 1.2 Photo- interpretation keys 

Aerial photo- interpretation is best carried out if interpretation 
keys are available. They are "reference material designed to 
facilitate rapid and accurate identification and determination of the 
significance of objects by the photo- interpreter" (ASP, 1960) . The 
purpose of interpretation keys is to provide the analyst with reference 
sources providing background information. These sources may consist of 
vertical and oblique aerial photography, ground photographs, and could 
be single or mounted in stereogramme form. 

Photo- interpretation keys can be selective or dichotomous. 
Selective keys are arranged in such a way that the interpreter selects 
from a set of examples which corresponds to the image he/she is trying 
to identify. This type of key is easy to construct and easy to use. 
In dichotomous keys (also called elimination keys) the interpreter 
follows a prescribed step-by-step process that leads to the elimination 
of all items but the one to be identified. The construction of 
dichotomous keys is rather difficult, especially when many classes are 
to be distinguished. Moreover this kind of key requires a precise 
standard description of the tree shape on aerial photographs which, in 
the case of mangrove forests, is not available at the present time. If 
established, dichotomous keys could however, be a valuable tool to be 
used to achieve a higher classification accuracy. 

A. 2. 1.3 Photo- interpretation aids 

Various enhancement techniques exist, which may aide in the 
analog (visual) interpretaion of aerial photographs, such as density 
slicing and the additive colour viewer. For details of these 
techniques please refer to Box A* 2.1 on page 271. 

The generation of digital data from a film transparency using a 
densitometer or an image digitizer can also be performed. However, 
this type of digital image data is generally considered of inferior 
quality. Analyses of digital data can best be made on multi-spectral 
imagery which is designed for such applications. 


Prom colour infrared video images, one can generate false colour 
composites, similar to those produced from satellite MSS imagery (Meisner and 
Lindstrom, 1985) . The analysis of video imagery is currently based on viewing 
the tape on a monitor screen using the still -image capability of the system, 
or simply on a hard copy photographed from the CRT. The image from the CRT 
may also be transferred to a map through simple instruments such as a Zoom 
Transfer Scope. The option of manually digitizing and interpreting the image 
is also possible by means of an image analyzer. No application of such a 
system to mangrove areas is known to the author. 


Satellite data are presented either in the form of images or in a 
computer compatible tape format. Information is extracted from images by 
analog (visual) interpretation while digital data analyses - called numeric 
interpretation - require the utilization of an appropriate device called the 
image analysis system. 

Satellite imagery offers the possibility to provide a quick and 
relatively cheap coverage over extensive areas. Because of the urgency of 
information needs, this kind of remote sensing is receiving much interest in 
vegetation study applications. In conducting surveys of mangrove vegetation, 
satellite remote sensing techniques (mostly Landsat) have been found to be 
highly effective. For mapping purposes satellite images can provide sufficient 
information for an overall planning at national and regional levels. 

A. 2. 3.1 Analog interpretation of satellite images 

Analog interpretation is concerned with a direct visual analysis 
of satellite images produced on paper prints or transparencies. Analog 
analysis has the major advantage of producing results (maps) which can 
be readily usable in the field. Moreover, a visual interpretation of 
satellite images can be achieved with a small input of personnel and 
materials. In order to achieve reliable results, visual interpretation 

Skillful interpreters; 

The quality of the interpretation depends on the interpreter who 
must be able to differentiate between colours and tones. He or she is 
also required to have a solid field experience for a rapid and 
efficient interpretation. 

Good quality images; 

Visual interpretation is usually made on paper prints or 
transparencies. The detail and accuracy with which vegetation cover can 
be identified depends to a large extent on the quality of the image, 
characterized by the degree of contrast and sharpness and the range of 
grey levels or colours. If the interpretation involves several scenes, 
tonal and spectral characteristics should be uniform. 


uflrt for aerial ittfttsoe^^ 

Density slicing ' " ' -" <'::.;:- - ; ;. : ...^; ....-. 

In this $>*oeess, which is used to afc*h<:& ^ 
image tones, different colours are assigned to v*ribu de&Sity 
allowing tonal diffeitettcies to be more mjRpat^fet; A topical 
slicing includes a light table, a video camera and a wlctar 

Additive colour viewing 

The reproduction of images on transparencies 
usually produces sharper images and a better definition than on paper 
prints. More details are discernable when transparencies are seen on 
a light tatole* In this form, black and white image tra^i^a^ettclietf can 
furthermore be observed with devices such as a colour adfi^Iti^ v&ew*r. 
The manipulation of various combinations of filters built into the 
device allows the analyst to determine - by trial and error - the best 
output suitable for feature separation. 


False colour composite generation \ 

In order to produce false colour composites, HSS bands are 
superimposed and a colour hue is assigned to each band. False colour 
composition can also be generated directly from the standard coti^mter 
compatible tapes using a film recorder of the colour VIZIR type. The 
interpretation of such colour composite is similar to that of infra red 
images (Borel r 1983), The author also pointed out that colour 
composites produced better results when the two first spectral bands 
are de-correlated. 

Diato colour film 

Another alternative which can be used to produce image 
enhancements is diazo colour film. The process of obtaining dlaso 
colour films has been discussed by several authots (Moore, 197$; Sfcaley 
et al,, 1977). With this procedure, a wide range of colours can be 
obtained at a relatively low cost by various combinations of 
and negatives of single bands. ^ 

Photographic contrast enhancement 

technique is performed to increase iittage 
by increasing density differences between scene feafctirfes, ' : ' : .':$t " 
mainly in stretching the density ran$e of the image over t 
density range of the film. One may also generate enhanced colour 
eotpoaites if single black und whit* b tt 
contrast -eflhawced before compositing them. 

Box A J.I: Enhancement techniques for analog photo-interpreUtkm 


In tropical areas, clouds can be a major problem as it may not be 
possible to obtain cloud free coverage of large areas at the same time. 
Images from different satellite passes can be acquired but one should 
be aware of the variation in spectral characteristics which may result 
in differences in tonal values and colour hues. 

To achieve an effective interpretation of images, it is necessary 
to pay much attention to the image characteristics of the different 
images. Some of the points to be considered are: the cloud cover, the 
spectral bands, the season of the year, and the scale of the image. 

Interpretation technique! 

Satellite images can be viewed in the form of single spectral 
bands but combinations of them can also be advantageously analyzed 
under a mirror stereoscope for a more detailed distinction (Baltaxe, 
1980) . Additional improvements in image interpretation can also be 
achieved using various procedures including density slicing, additive 
colour viewing, false colour composite generation, diazo colour film 
and photographic contrast enhancement. For further information on 
these techniques refer to Box A* 2.1. 

Auxiliary data 

During the process of image analysis, base maps - if available - 
are of utmost importance because they may provide accurate information 
on the location and identification of land features. Their 
incorporation is made easier when their scale is similar to that of the 
images used in the interpretation. 

In vegetation studies, the introduction of auxiliary data 
contained in topographic or thematic maps was found to be very valuable 
for detail recognition and delineation. The procedure consists of 
superimposing the maps and positive transparencies on satellite imagery 
of the same scale. Planimetric details from a topographic map for 
instance can be transferred to a plastic overlay. This overlay is then 
superimposed to the image and land-use pattern and boundaries of each 
type are simply delineated by tracing over the transparency. The 
inclusion of auxiliary information may ensure a decrease in costs 
incurred in ground surveys (checking the classification) and may result 
in an overall reduction in interpretation time. 

In the process of image analysis, the addition of secondary data 
from maps is not limited to visual interpretation. It can equally be 
used in studies where small format photography or digital data analysis 
is applied. 

IquipMnt ttMdtd 

Visual interpretation of satellite imagery does not require more 
than what is needed for conventional aerial photography interpretation. 

Basic equipment includes: 

Light table (with uniform illumination to ease the strain on the 
eyes) ; 

Transparent paper. In order not to damage the image, delineation 
is best carried out on paper which is super -imposed; 

Magnifying glass; large low power lens; 

Dot grid for area assessment; 

Drawing pencils and pens; 

Tape ; 

An additive colour viewer and a mirror stereoscope are also 
essential instruments which can be of great value to the 

A. 2. 3. 2 Computer-aided classification 

Digital data which is recorded on computer compatible ta>es (CCT) 
must be processed on an image analysis system. The physical aspects of 
such systems are out of the scope of the present paper but it might be 
indicated that a typical system is composed of a computer, a terminal, 
a colour display monitor and other peripherals such as a printer, a 
plotter and a digitizing table. To successfully operate an image 
analysis system, some experience in computer utilization and 
familiarity with digital classification are required. 

The process of digital data analysis passes through three phases: 
preprocessing, enhancement and classification. Digital data 
preprocessing and enhancement are generally conducted to improve the 
interpretability of images. They can be applied to virtually all kinds 
of MSS data. The interpretation which follows preprocessing and 
enhancement can then be automatic (computer classification) or visual 
from a screen or a photographic image. 


Before any classification begins, radiometric and geometric 
corrections of the MSS data should be undertaken in order to eliminate 
the various anomalies and defects which may have occurred during the 
process of data recording. Anomalies are caused by transmission losses 
or sensor saturation, and may result in missing scanning lines and/or 
in lines of contrasting brightness across the scene. Such defects can 
be eliminated by a radiometric correction. In the case of a missing 
line, pixel values are usually replaced by the average of the pixels 
immediately above and below them. This correction should only be 
applied in cases where no more than a few missing lines are 


Atmospheric scattering of short wave lengths can also cause a 
reduction in contrast* In this case, it is suggested to apply a 
subtraction of a constant value from all spectral radiance quantities. 

Image distortion is another problem. This is caused by the 
earth's rotation, sensor delays and orbit altitude variations. 
Therefore, it is necessary to apply a geometric correction to adjust 
the spatial position of each pixel in the scene. 

BnhanctMnt techniques 

In the same manner as analog image enhancement, there are also 
digital enhancement techniques which can be employed to facilitate 
image interpretation. The main procedures involve contrast enhancement 
and data transformations. Please refer to Box A. 2. 2 for details. 

Image classification 

The primary goal of digital data processing is to extract useful 
information from the image. Having this in mind, the analyst must 
select the procedures or techniques which allow him or her to bring out 
the image features judged to be most important. 

Through one or all of the enhancement techniques described above, 
the computer is used to make it easier to interpret certain features of 
interest in a more detailed and accurate way, but the interpretation 
can also be performed by visual means. 

Prior to the classification itself, it is always useful to 
conduct some sampling analyses in order to obtain insight into the 
interpret ability of the data relevant to the categories to be 
differentiated. One may examine the frequency distribution of spectral 
values in each band and/or analyse feature space plots which can be 
generated by plotting radiance values in a given spectral band against 
those of other bands. 

The two basic approaches which are employed in digital 
classification are supervised and unsupervised classifications. The 
supervised classification utilizes "training 11 information obtained from 
sample sets of spectral clusters which correspond to land-use or land 
cover classes that the interpreter wishes to distinguish. The sample 
sets should encompass the spectral diversity in the class. They must 
therefore be spread throughout the image. The choice of spectral 
classes and their number are key elements for the success of the 
classification. Furthermore, they must be distinct enough to assure a 
reasonable accuracy. Once the spectral classes have been defined, the 
second step in supervised classification is to classify all the image 
pixels according to the spectral characteristics obtained from the 
training areas. Based on this information the -classifier places 
unknown features in the most likely group. Among all classifiers in 
use, numerous studies have shown that the maximum likelihood 
discriminant analysis provides the most accurate classification. 


Hie ofcjctl<v of contrast ^iemiit i* fee 

interest to be separated using tonal values. For this, purpose, the 
frequency distribution histogram of spectral rlihc* valiww if VMfMl$ 
(histogram stretching) by expanding the range of the pixel values of a 
given area of interest over the entire range of the display device, 
Histogram equalisation can also be performed in order to n*are thst the 
image levels fit the distribution of digital values. It consists of 
assigning equal numbers of pixels to each tone level. 

Edge enhancement is another technique which is applied to increase 
tonal separability between features which exhibit small viiiations in 
tonal values along their edges. With regard to mangroves, th is 
technique may be useful in drainage network delineation. / 


Modifications of digital data are f recently tnade thurougfh various 
transformations of spectral values. These trat onwtions include fc*n& 
ratioing, canonical and principal components, band de-correlation and 
image smoothing. 

Bead ratioingi 

Thi. technique i* oonroonly applied to M3S dmtm. It con*i*t* P dividing th 
epeotral value* in one band by the value* in another band on m pixel fey pixel ftteitf all 
over the aeeae. ftatio* of the difference to the man of band* and ratio* of each bend to 
the um of all the hand* *re commonly applied to prodms* derived data etn. Fpr 
vegetation atudie*. the ratio of band 7 /band 5 of Landaat, termed vegetation index (w 
it oorrelatee with green vegetation), hae been frequently ueed a an index of green 
hiottaae. Mulder and Venqveniua (1^74) indicated that the application of ratioing 
technique* reduce* the variation in brightnea* on wulti- temporal *cene 

Prineipal oa*ponenti 

With principal oow^onent techni^uee, the epecferml vmluw txm all hand* are 
cororeeted in the form of a *et of un^correlmted axia in order to mi^citaiie the variaj>oe 
between claeeee (categories to he identified) end miniftite it within each Claw, fltie 
procedure i. vmlumble in digital 4ata amlynie ^Aien the iNttfprete>r i 4pUto;*U& * : 
large number of peetral band* and ita application le effective without eignifioant loe* 
of information ' 

Xaage Moothiag (apatial f ilterija) - 

Image smoothing i a data tramformation in which block* of **3 or Sx5 pixele are 
paed through a apatial filter in order to remove eome : e t}a high ^f 
in the data. In the proce*., the digitel vattif of the **JWa flM4, 
{window) of the eoe^e, i ^wleoea by th* nveM *f )tal *4*1 j*1^ C 
constituting the block. *ometiee, the median ie u*ed *Wt*^4. *$ ttm 
applioatio^of thi* technique to Und*at data hae .been shown to be efficient 
fcre*t type boundarie* (Pox, 



blue and green band* either from I*nd*at or *05t ere 

u MMIIIHII w "^^ '- T-"~- 7T7 -^ , . -^ "^- ,-,^ A .. 

coatrwt ffct vh-an colour o<poitM ar madto 




Conversely, the untuptrviMd classification uses only the 
statistics of the data on which the classification is based. In this 
way, pixels from samples areas are examined in order to determine 
natural groups of data based only on their spectral properties. It is 
then up to the analyst to assign meaningful classes to the clusters. 
The unsupervised approach is practical when the number of classes is 
high which, would make the process of using training areas very 
tedious . 

The combination of both supervised and unsupervised strategies 
has however been found to produce the best results. 


The Synthetic Aperture Radar (SAR) sensor follows the same basic 
principles whether it is mounted on board an aircraft or a satellite. This 
sensor, operating in the microwave portion of the electromagnetic spectrum, 
has been used for vegetation studies in many tropical forest areas. Most of 
them were based on aircraft mounted SAR. 

Radar sensors do not generate images according to the human eye. One 
should be aware that some forest types are different in tone on SAR images 
compared to aerial photographs. Moreover, differences in tone on SAR do not 
always correspond to differences in vegetation, and important differences in 
forest types are not always reflected in differences in tones on SAR. Because 
of this, a pre-established legend has no significant value in Radar image 
interpretation. It is therefore suggested to: 

Separate roughly between forest and non- forest areas; 

Use physiographic features to delineate broad forest types; 

Use additional data from reconnaissance surveys to finalize vegetation 

SAR imagery is not as well known as photographic or MSS imagery, mainly 
because it only recently became available to civilian use. In fact, with a 
few exceptions, its applications to earth resources is still at a research 
phase. But, future plans to include SAR in the pay load of orbiting 
satellites such as ERS-l (European satellite) , JERS-l (Japanese satellite) and 
others, will open up new perspectives of more frequent applications of this 
system. Like MSS data, the new developments in image data processing can also 
be applied to SAR imagery which can be produced either optically or digitally. 

In addition, since SAR senses micro-relief on surface configuration and 
dielectric properties of vegetation, its combination with MSS satellite 
imagery can increase discrimination and identification of vegetation 
boundaries (Gelnet et al., 1978). 

The results obtained from different vegetation studies where SAR data 
been used, indicate that research has only just begun to explore the 
techniques of interpreting Radar imagery. However, because of its complexity, 
the interpretation and analysis of SAR images necessitates well trained image 
analysts, who are acquainted with preprocessing, postprocessing and 
enhancement procedures, specific to the system. 



Maps play an important role in forest activities. The information to 
be contained on a map depends on the objectives for which the map will be 
used, and the scale of the map, which is determined by the level of 
application. This application can be national, regional, provincial or even 
a district or a smaller unit level. 


A. 3. 1.1 Maps from aerial photography 

Map compilation consists of producing a document - based on 
photo- interpretation - from which the final map can be prepared for 
printing. Before the procedure starts, care should be taken that no 
doubtful lines are drawn and the photographs are free of superfluous 
information. Moreover, it is important to check that the class 
delineations from one photo to the next are matching. This provision 
facilitates the transfer of cover classes to the base map and prevents 
editing errors. 

If delineation is carried out on transparent sheets, these must 
carry fiducial marks and pass points to facilitate their orientation. 
It may also be useful to match photocopies of delineated transparent 
sheets to obtain a rough map which can be used to check the photo- 
interpretation (Remeijn, 1985) . 

Base map 

Details of each land-use class, forest type and other strata 
which have been identified and interpreted on aerial photography must 
be transferred onto a base map. The function of a base map is to 
indicate the geographical location and distribution of the terrestrial 
features plotted on it, therefore, it should contain some topographical 

A topographic map is usually adopted as a base map. If used as 
it is, however, the addition of more information might produce a map 
with too many superfluous details, which may hamper readability. It is 
therefore suggested to simplify it to an extent that it still gives 
sufficient information to allow the map user to locate in the field the 
features drawn on the map. Items that could usefully be included 'are 
roads, rivers, urban areas, etc. Contour lines are generally not 
necessary. Over relatively flat areas, such as mangroves, their 
density is in any case not very high. 

Aerial photographs can be used to provide topographical 
information for the base map if no topographic maps are available, or 
existing maps are of too small a scale to provide reliable feature 
location. Aerial photos can also be employed if terrain features 
identifiable both on the map and on the ground are lacking, or existing 
maps are outdated. The photos should in this case include a number of 
control points to ensure mapping accuracy. For that purpose, a radial 
triangulation is carried out. 


This gives the coordinates (x,y) of control points which have 
been identified in the field or from geodetic sources. Slotted 
template and graphical methods are simple to use and can provide a 
satisfactory network or supplement an existing network of control 

For nearly flat terrains, it may be possible - and economical too 

- to use every second photograph both for interpretation and control 
point determinations. 

Transfer instruments 

The transfer of photo delineation onto base maps can be performed 
by means of various transfer devices. The transfer can be achieved by 
either stereo or mono instruments with varying capabilities of 
adjustments for image distortion and scale differences between 
photographs and the map. 

The instruments which are generally used for transferring details 
from interpreted aerial photographs to a map vary in performance, 
precision, handiness and cost. For some cartographic tasks, including 
mapping for forest surveys, the accuracy requirements do not justify 
the use of sophisticated and therefore expensive devices. In mangrove 
areas, characterized by relatively flat terrain, simple methods and 
equipment may be adopted to produce satisfactory results. Most 
advantages of simple devices reside in their low cost, easy 
transportation, simple construction and operation, and - more relevant 

- their simple maintenance. A few of the simple instruments are 
presented in Table A. 3.1. 

Table A J.I: Examples of simple transfer instruments 




OMI Stereo facet plotter 

Reflecting projector 

(optical pantograph) 

Bausch and Lomb ZTS 





mono- stereo 

Technical characteristics, construction principles and 
performances of these instruments can be found in various documents and 
textbooks (Weir, 1961, 1983). The instruments presented above are 
suitable for the transfer of details of photographs of flat terrain, 
where displacement errors do not occur or are negligible, and where the 
main source of error is the camera tilt at the moment of exposure. 

In spite of their advantages in terms of cost and handling, it is 
important to note that simple instruments provide only a partial 
elimination of geometric errors inherent to the aerial photograph. 


Where high planimetric accuracy is required, one has to resort to 
stereoplotting instruments. These devices are also capable of 
producing planimetric details and contours directly from stereoscopic 
models. In addition, modern stereoplotters are equipped with 
coordinate readouts which are more convenient to use with data in 
digital form. 

A. 3. 1.2 Maps from satellite imagery 

At a small scale, which characterizes satellite images, the 
objective of image interpretation is primarily to produce maps showing 
mangrove forests along the coast and rivers. The information which can 
be obtained from the images (although it is only at reconnaissance 
level) , is very useful for a forester or land manager in areas where 
basic information is lacking. The map and other data obtained from the 
image will give an overall indication of the actual land-use pattern, 
which is sufficient for planning at a national or regional level. 

Maps can be directly drawn either from photographic prints or 
from computer-aided interpretation outputs. 

Mapping from photographic images 

Mapping from Landsat imagery, which has been commonly applied, is 
usually carried out with black and white prints of single spectral 
bands 5 and 7, at a scale of 1:1 000 000. For a more detailed 
interpretation, colour and colour prints or transparencies and 
enlargements to scales of 1:500 000 and 1:250 000 are also frequently 
used.. At these larger scales, the observation of details and their 
delineation is made easier although no increase in resolution is 
obtained. From images produced by other more recently launched 
satellites, such as Landsat 5 and SPOT, equipped with high resolution 
sensors, enlargements to 1:100 000 and 1:50 000 can be obtained without 
much loss of definition. 

On satellite images, the position of ground points is subjected 
to geometric errors due to various factors such as the satellite 
altitude, the orbit variation and earth's rotation. A correction of 
these geometric errors can be achieved using a number of ground control 
points. But, unlike aerial photographs, which contain geometric 
distortions requiring adequate transfer devices, satellite images 
exhibit only small geometric discrepancies when compared to topographic 
maps of the same scale (Wong, 1975) . For this reason, standard 
satellite images are generally considered to have sufficient accuracy 
required for forest survey applications (Baltaxe, 1980) . 

The simplest procedure of mapping from satellite images consists 
of delineating boundaries of vegetation cover and other land uses 
directly on the image. Image interpretation can be transferred simply 
by tracing photo details onto the map, the latter being prepared on 
transparent material. On the image, the interpretation way 
alternatively be carried out under a stereoscope, and feature 
delineation can simultaneously be drawn on a transparent sheet placed 
on one of the scenes. Later on, each transparent sheet, associated 
with each scene, can be adjusted according to the base map, on which 
ground control points and other topographic features have been plotted. 

Napping froai automatic intarpratation 

When computer facilities are available, and image data are used 
in digital form, classification results are often produced by a line 
printer. The output is thus a digital map where each single pixel is 
represented by a symbol, a letter or a digit, chosen by the image 
analyst. Also, various grey levels during printing can be obtained so 
as to facilitate the map readability. This is in fact just a matter of 
software . 

This kind of output can also be produced in colours, providing 
multi-colour printers are available. Due to printer sizes, and since 
each image pixel is represented by a printer character, it is usually 
necessary to assemble many computer outputs to cover a whole scene. 
Compressed outputs may be produced to accommodate large area coverage 
and reduce the number of sheets which in practice pose some handling 
difficulties. Nevertheless, for a provisional analysis, this type of 
maps may be useful. More condensed versions of computer maps can also 
be produced on dot matrix and ink- jet printers, which are able to 
generate a wide range of grey levels and colours. 

Photographs, taken directly from the monitor screen, on which the 
classification is displayed, have also been used as map products. 
Obviously, image distortions caused by the curvature of the screen do 
not allow correct measurements to be made on the photos, but areas can 
still be determined based on pixel counts. 

Higher quality products are obtained by means of appropriate 
devices which convert computer processed classification into a hard 
copy image, generally on photographic material. Due to the high cost 
of equipment, computer processed data can be recorded on tapes and sent 
to specialized centers for their conversion to photographic films 
either in black and white or as colour products. 

A. 3. 1.3 Mao accuracy 

When a map is produced, it should be evaluated for its accuracy. 
Two kinds of accuracy can be distinguished: geometric and thematic 

Gaoautric accuracy 

Geometric accuracy concerns planimetric and altimetric accuracy. 
It is generally expressed in terms of the precision in the geographic 
position of map features, given a certain range of tolerance. Geometric 
accuracy is a function of the base map construction, the procedure used 
to determine the coordinates of ground control points, and the method 
and instruments used in the transfer of photographic details unto the 
base map. 

For forest surveys, photographs and the maps derived from them 
constitute the base for measurements and planning operations. In order 
to achieve reliable inventory results, reasonable planimetric accuracy 
should be present in forest maps, particularly those used on the forest 
management and operational level. 


Boundaries must be properly defined, and vegetation types 
correctly indicated. This is of utmost importance when property and 
concession boundaries are considered. Planimetric accuracy affects to 
a large extent the area measurement which, together with the volume per 
unit area, determine the total timber volume of a tract, a concession, 
a stand or a larger unit. 

Planiattric accuracy can be determined by means of measurements 
of a number of distances on the map and the ground or other reliable 
data support (topographical maps for instance) . It may be pointed out 
in this respect that planimetric accuracy of a map is also dependent on 
the stability of the material on which the map is produced. In 
tropical conditions, characterized by high temperatures and humidity, 
this is an important criterion to be taken into account in the choice 
of map supporting material. 

Altiaatric accuracy is of little importance in mangrove areas due 
to the flat terrain. 

Thematic accuracy 

In recent years, the development of new techniques for collection 
and processing of remotely sensed data has progressed rapidly, allowing 
for mapping at low cost and with different output formats. Given the 
cost advantages, the question posed is whether accuracy is sufficient 
when remote sensing imagery is the principal data source. 

Thematic accuracy, which is the accuracy of interpretation, is 
associated with the ability of the interpreter to correctly identify 
terrestrial features from the images. Thematic accuracy is thus a 
function of image characteristics such as print quality and scale but 
is also dependent on the skill and experience of the interpreter. 

When producing maps to be used at the forest management and 
operational level, the accuracy of the photo delineation is associated 
with the ability of the interpreter to distinguish between different 
types of vegetation, different species, different stand age and stand 
site quality. The reliability of photo based classification has been 
addressed in many studies. Ground verification seem's to be a logical 
answer but due to time and cost constraints, a valid sampling procedure 
is required. Several techniques have been used - both qualitative and 
quantitative approaches. 

The accuracy of image interpretation can be assessed from a 
sample of points on which cover classification is confronted with the 
ground truth. Misclassification affects area determination, which must 
be adjusted after such field checks. 

It should be noted however, that misinterpretation can be reduced 
in many instances by careful selection and good training of 
interpreters and continuous checking of results together with limiting 
the stratification to a reasonable number of easily determined strata. 


For extensive areas, where the interpretation involves small 
scale satellite data, aerial photography at medium and large scale can 
be used to complement field checks. Small format camera systems 
mounted on a light aircraft may in this respect turn out to be a 
suitable solution as conventional aerial photographs may be outdated or 
not available. At low altitudes, sample strips can be taken to 
complement the costly ground checks, especially in remote areas of 
difficult access. 

The sampling problem is one of determining the optimum number of 
map sampling points to be compared with ground points and the sampling 
design to be used. A random selection of points based on the binomial 
distribution was suggested by Hord and Bruner (1976) . In order to 
satisfactorily represent the smaller areas in the sample, stratified 
random sampling is recommended. Rosenfield et al. (1982) and Hay (1979) 
also indicated that stratified sampling should be applied with at least 
50 observations per map category. The procedure consists, of sampling 
each category, either separately or in combination with the others. In 
the latter case, a sample is drawn randomly over the whole area of 
interest and the number of points falling in each stratum is cumulated 
until each one of the strata has a sufficient sample size. During the 
process of sampling, samples which fall in a category which is already 
filled are rejected, whereas those falling in other categories are 
retained until all strata are completely sampled. 

Samples can as well be systematically distributed over the whole 
area using a regular grid. With a regular distribution of sample 
points, the number of samples in each category will be proportional to 
the stratum size. Care should however be taken that small strata are 
sufficiently represented in the sample. This may cause some problems 
if the grid has a low density and/or map categories are broken up into 
small patches. 

Whatever the sampling design used, the actual classification is 
compared to the 'correct' classification and the accuracy is expressed 
in terms of the percentage of sample points correctly classified. 

Digital classification of aerial photography and satellite 
produced imagery has also been investigated for its accuracy. When 
compared to the manual selection of sample points on the map, a 
computer based sampling procedure provides more adequate samples of all 
categories and results in easier analysis (Fitzpatrick-Lins, 1981)'. 
For digital classification, the discrete multivariate analysis 
technique is usually applied to assess accuracy. The method consists 
of constructing an error matrix, also called a confusion matrix, where 
row and column elements correspond to the number of cells or pixels 
classified on imagery and on the ground (or other ground truth source) 

Table A. 3. 2 below is an illustration of such a confusion matrix. 


Table AJ.2: An example of an error matrix 




1 2 r 



x n x, a x lr 




X 2J 






















X r j X rr 

x r . 


X+j X+2 ***r 

N i 

In this matrix, diagonal elements represent the number of units 
correctly classified. The method, which has been commonly applied to 
the accuracy assessment of satellite data classification, is based on 
a measure of agreement of square matrices presented by Bishop et al 
(1975), and which was extensively discussed by Cohen (I960, 1968), and 
Fleiss et al (1969) . 

For a multi-nominal sampling model where the total number of 
elements is fixed, an estimate of the agreement coefficient is given 

r r 
N Ex,, - Ex,.x^ 


N 2 - x l+ x 41 


k coefficient of agreement 

Ex tl - sum of all diagonal elements 

x i4 = marginal total of row i 

x +1 - marginal total of column i 

N * total number of units in the sample 

r * number of categories (strata) 

The classification accuracy of a given class i is obtained 
by dividing the marginal total in row i by the marginal total of 
column i (xi+/x+i) , and the overall accuracy is simply computed by 
dividing the sum of diagonal elements by the total number of points 

The coefficient of agreement can also be employed to assess 
the photo classification accuracy for each cover type or all cover 
types simultaneously. 


Congalton et al. (1982) found this procedure to be useful 
in evaluating photo -interpret at ion accuracy and in comparing photo- 
interpreters performances. The comparison is based on a statistic 
test which can be developed to determine if there is a significant 
difference between two independent K's. This test is achieved by 
evaluating the normal curve deviate, since K is asymptotically 
normal (Bishop et al, 1975). 

Instead of point counts, the error matrix may be filled 
with areas determined from the image classification and from a 
reference data source. Then, one may compute correctly classified 
areas, omission and commission errors and overall accuracy (see 
Table A. 3. 3). Lantieri (1986) suggested that the computation of 
classification accuracy should take into account the commission and 
the omission errors in each class, as it is defined below. 

Table A 33: An example of an error matrix with area Tallies instead of point counts 


S(i): Xrea correctly classified 

C(i) : Commission error on class i: Area classified in i on image but is not class i in the terrain 

0(i) : Omission error on class i: Area in class i on the terrain but not classified as i on the image 

8(i)+O(i): Total area of class i on the terrain 

PC(i): Mapping precision of class i (in %) : [*(i)/8(i)+0(i)+C(i)]100 

P8(i): Statistical evaluation of class i (in %) : [8(i)+C(i)]/(fi(i)+0(i)]loO 

Pd(i): Height of class i among n classes : [S(i)+0(i)]/d(8(i)+0(i)] 

PCT: Total mapping precision (n classes): d Pd(i) PC(i) 

(Source: Lantieri, 1986) 


In some instances, where no adequate maps are available or their 
compilation is lengthy, photo mosaics can be used as substitutes. In a photo 
mosaic, images are assembled and fitted together to form a document which 
gives an overall view of the area being covered by aerial photography or 
satellite images. 

The simplest form of a mosaic is a photo lay-out, where prints are 
mounted together to obtain the best fit of image detail. In this form, the 
mosaic - named uncontrolled mosaic - will have appreciable planimetric errors 
caused by camera tilt, scale variation and relief displacement distortions 
contained in the original prints. For field survey operations and navigation, 
uncontrolled mosaics are however useful. Errors can be reduced by using the 
central part (the effective area) of each photograph. 

If measurements are to be taken on the photo mosaic, the latter 
should have a reasonably consistent planimetric accuracy. One should 
therefore resort to a controlled mosaic. For its construction, photographs 
should be brought to a common scale and the effect of camera tilt removed. 
In mangrove areas, where relief displacement is negligible, photo distortion 
due to camera tilt can be eliminated by means of optical rectification. The 
construction of a controlled mosaic requires the determination of a number of 
ground control points which can be obtained from an aerial triangulation. The 
result is a photomap where terrain features are represented by their 
photographic image in their true planimetric location. It is worth noting 
that the cost of mosaic compilation can be reduced if small scale aerial 
photographs are used both for triangulation and rectification before they are 

For a more effective use of the data, strata delineation on aerial 
photography can be transferred onto the photomap, by means of adequate 
instruments (generally, a pocket stereoscope is sufficient in case of 
homogeneous scale) . On the photomap lineal and area measurements can directly 
be made for planning management operations, timber sales etc. 

Satellite images can also be employed for mosaic construction, and 
where no aerial photography is available, such mosaics are extremely useful. 
Good quality images should however be selected in order to ensure a uniform 
appearance in terms of brightness and tonal values. Complementary photographic 
treatments might be required to obtain more homogeneous tones on images of 
different passes. Moreover, marginal information must be used to obtain a 
good detail fit and correct position of scenes with respect to each other. 




The objective of sampling is to gather reliable information at low cost. 
In view of the problems posed by accessibility and working conditions in 
mangroves, it is strongly recommended, in designing a survey, to incorporate 
any element that may contribute to increasing the accuracy of forest 
classification while at the same time making the field enumeration less time 
consuming. Low altitude aerial reconnaissance, stratification and other 
sampling procedures such as multi -phase and cluster sampling are some 
approaches which deserve consideration. The description of these designs and 
the associated statistical formulas are extensively discussed elsewhere 
(Lanly, 1973 and others) , and are thus only briefly summarized below. 


Stratified sampling designs are frequently applied in surveys and forest 
inventories. The object of stratification is to subdivide the forest into 
more homogeneous parts, in order to reduce the variability of the parameter 
to be estimated. 

The allocation of sampling units to strata can be proportional to the 
strata area or to the variance of the strata. These approaches are discussed 
in various forest inventory documents. The optimum allocation approach 
requires advance estimates of the variation in each stratum, which can be 
provided from pilot surveys or, if available, from past surveys and 
inventories in similar areas. 


Basically, the procedure involves the selection of large units in the 
first phase, named primary sampling units. Within each primary, a number of 
smaller units - secondary units - is drawn. The procedure can have more than 
two phases and can use varying methods of selection in each phase. 

Its disadvantage however, is that the concentration of the ultimate 
sampling units results in a larger variance of the estimate compared with a 
one phase design of the same sampling intensity. 

A. 4. 2.1 Double -phase sampling 

The application of doublt-phaM sampling for stratification in a 
forest survey results in an improvement in the stand characteristic 
estimate through a better estimate of strata areas. 

In this method, a large number n of photo plots are drawn in the 
first phase, and a subsample n' of the first sample is selected in the 
second phase to be used in the field. The main objective of the survey 
is to provide an estimate of strata proportions in the first phase, 
based on some stratification rule 1 allowing the first phase sampling 
units to be classified into land-use classes, vegetation cover classes 
or other classification criteria. The second phase plots are used to 
check the photo classification and collect data on forest and tree 
characteristics . 


Advantages of this technique have been found to be substantial in 
surveys of large areas. The design is more complex than a single stage 
random sampling but it is more efficient. 

The application of this design usually involves aerial 
photography in the first phase. Satellite imagery can also be used 
but, because of its low resolution, it might be very tedious or even 
impossible to correctly locate the subsample units on the ground. 

In non- forest classes and areas where the forest has been so 
severely disturbed, that no commercially valuable trees exist or the 
forest is poorly stocked, data collection may be restricted to simple 
observations on vegetation status, 

Areas are estimated from the first large sample of photo plots. 
The proportion of each stratum is estimated by : 

P b -- 

n h is the number of photo plots falling in stratum h and 

n is the total number of photo plots in the first large sample. 

The area Ah of stratum h is estimated by: 

A. = P h *A 

Where A is the total area concerned by the survey, assumed to be 
known . 

The information collected from the ground plots is used to 
correct biases in area estimates caused by various sources of 
misinterpretation. The adjustment is applied to strata proportion in 
the following manner: 

The adjusted stratum proportion P h is obtained by: 


Adjusted P h - E P,*PH 

M is the number of strata and 

n h 

with n hj being the number of ground plots actually falling in stratum 
j, but classified h on photos and 

n h being the number of photo plots classified in h. 


The variance of the adjusted stratum area proportion is computed 
by the expression: 

N P/p^d-py) 1 M M 

V(Adj. PJ - E ............ + - [EPjp hj a - 

n h n 

which is a simplified formula for double sampling for stratification 
with a discrete random variable having attribute 1 or 0, and where the 
term P h /n has been dropped being considered negligible. 

Data obtained from ground sampling units can also be used to 
estimate mean values of stand characteristics such as timber volume, 
stocking, etc. The estimate of the mean value per unit over the whole 
area concerned by the survey, is given by the expression (Lanly, 1973) : 

_ MM 

y. t = E E 

where y h3 is the mean value per sampling unit of the 
characteristic y in the part of plots actually in stratum j, classified 
in stratum h by photo- interpretation. 

The mean value per unit in stratum j is estimated by: 


The estimate of the total Y over the whole area inventoried is 
obtained by multiplying the estimate of the overall mean y. t by the 
total area A, and the total in stratum j is computed by multiplying the 
mean of stratum j by the term A(EP,p h j) which is the corrected area for 
stratum j . 

Double -phase Mapling for regression is another technique which 
involves two variables, the main (y) and the auxiliary (x) . It is a 
powerful procedure which is frequently used in forest inventory 
sampling. It is particularly useful when the cost of enumeration of 
the main character is much higher than the cost of the auxiliary 
variable, the latter being correlated to the first one. The approach 
is recommended when the inventory can make use of both aerial 
photographs and field enumeration. 

In the first phase, a large sample n of photo plots is drawn from 
the population N. The stand characteristic of interest (represented by 
the auxiliary variable x) is first measured on photo plots. This can 
be a gross volume estimate for example, based on measurements of the 
stand height or the crown density on the photos. In the second phase, 
a subsample n' of the first phase large sample is taken and 
measurements are made on both x and y. y may well in this case be the 
volume per plot, which is determined in the field through conventional 
techniques, while x is ground measurements of either stand height or 
crown density. 


Double sampling for regression is also used in inventories on 
successive occasions. It may involve completely independent samples, 
or, in case of permanent (CFI) plots, it uses subsamples of the 
original sample or samples which are partly independent and partly 
subsamples. The latter case is termed sampling with partial 
replacement. In either case, change evaluation is determined through 
regression analysis between measurements made on successive occasions. 
The technique involves rather complex computations. Relevant 
estimators are presented in various forest inventory and statistical 
textbooks . 

A. 4. 2. 2 Three -phase sampling 

The design is similar to that of double-phase sampling for 
stratification except that more phases are considered. In the first 
phase a simple random or systematic sample n of large size is drawn 
from the population concerned by the forest survey or inventory, and 
sampling units are classified into predefined strata. From this first 
phase stratification, nl, n2,...n h , sampling units are obtained where 
n h in the number of units in stratum h. (h- 1,...L). 

The second phase consists of selecting a subsample n^ from n h . 
The selected n^ units of phase one are further stratified into the same 
or different strata as in the first phase one, and m^ second phase 
sampling units are obtained for each stratum j of second phase in each 
stratum h of phase one. 

In the third phase a subsample in each second phase stratum is 
drawn. The units selected in the third phase noted b h , are then used 
for measurements of the characteristics of interest. Observations are 
noted y h3k , where k= l,...b hj . 

A typical three-phase sampling design incorporates satellite 
imagery (phase one) , aerial photography (phase two) and ground sampling 
(phase three) . A four phase design can as well be employed if both 
small and large scale aerial photography are used in the second and the 
third phases respectively. 

One question which may be posed concerns the sample sizes in each 
phase. Theoretically, sample sizes in each level should be determined 
in such a way that the total survey cost is minimized. The problem 
becomes one of optimization which is out of the scope of the present 
discussion. For more details on the question and the computation of 
the variance of estimates, the papers of Frayer (1979) and Jeyaratnam 
et al. (1984) are recommended. More simply, sample sizes associated 
with each phase, can be defined arbitrarily before the first phase 
selection. It is, however, not recommended to use less than two 
sampling units per stratum. 

In the three-phase sampling case, the estimate of the total value 
of the population parameter is obtained by: 

N L n, J h 
Y - - E -- E 

n hl m,, jl 


The application of a three-phase sampling and the above formula 
to mangrove forests can be illustrated by the following example: Let 
N be the total number of sampling units contained in a given area to be 
inventoried* Let each sampling unit be associated with a pixel from 
Landsat imagery (say of 1:250 000 scale), used in the survey as the 
first level of sampling. We consider further that medium scale (1:30 
000) aerial photography of the same area is available and could be used 
in the second phase. The third phase is the sampling in the field 
where stand characteristic measurements are taken. The sequence of the 
method is described as follows: 

Step 1. 

From the total number N of sampling units, say 262 144, a sample 
n is drawn at random or systematically on satellite images. Out of a 
total of N units, n 13 107 (5 % sampling intensity) units are then 
classified according to defined land-use classes on satellite images. 
The number of different classes (L) will be a function of the sensor 
and classification procedure used. The latter can be computer-aided or 
visual. For sake of simplicity, let us assume that three classes 
(strata) are defined in this first phase: Forests, non- forest and 

The results in this first phase may be: 

Stratum number Land cover sample size 

h-1 forest n, 9 830 
h2 non- forest n, 1 966 
hL3 water n s 1 311 

En 13 107 

Step 2. 

In the second phase, samples are drawn from the first phase 
samples. Using some predefined sampling fraction (1% for instance), 
these phase two samples which are noted mj-983, mj-197 and m^lSl are 
located on aerial photographs and stratified according to a more 
detailed interpretation using photo variables such as crown density, 
tree height, etc, and other criteria on non-forest areas such as 
agricultural lands, salt ponds, fish ponds and shrimp farms. We assume 
in the example that the first phase stratum "forest' is refined into 
three strata in phase two, namely "dense", "open" and "degraded". The 
second stratum of phase one is split into three strata also. They are 
"agriculture", "aquaculture" and "other". 

Finally, we consider that the third stratum of phase one which is 
"water 91 remained unchanged in the second phase. According to this 
scheme, the results of classification are: 

Stratum number 



Non forest j2 




Land cover 




sample size 

m n -464 1 
m ja 197 \ 
rn^-322 J 


'maa-59 \ m, 197 
m a ,-38 J 

m^-131 m, 131 

Eml 311 

Step 3: 

In the third phase, samples are selected from phase two samples 
to be located in the field. Assuming again a sampling fraction of 1%, 
the final number of sampling units to be measured in the field will be: 

Stratum number 

Land cover 

sample size 

b bj 



b u -46 

2 760 




1 930 



b l3 -32 










b ja 4 



- b hj is the number of third phase samples drawn from the 
second phase samples m^. 

- y bjk is volume measured on the ground plots. 

The values in the last column are assumed to be the total values 
of timber volume in each third phase stratum expressed in m 3 . In non- 
forest classes and areas where the forest has been so severely 
disturbed, that no commercially valuable tree exists or the forest is 
poorly stocked, data collection in the field may be restricted to 
simple observations on vegetation status. 

The estimate of the total timber in the whole area concerned by 
the survey is given by the formula -presented above, which yields the 
following result: 


262 144 f 9 830 464 197 322 

Y | [ -(2 760) + --(1 930) + -(530)] 

13 107 [ 983 46 20 32 

1 966 100 59 1311 131 } 

+ [ --- (0) + -- (0) ] + ---- [ (0)] I 

197 10 6 131 13 J 

10 436 884.25 m 3 

Other parameters such as the total forest area may also be 
estimated using multi-phase sampling techniques. In that case, the 
variable y hjlc takes on the value 1 when the sampling unit falls in the 
stratum forest and zero otherwise. 

In a two-phase sampling design, following the same pattern as 
above, the estimate of the population total is given by: 

N L n h m h 
- E ~ E y hj 

n hl m, j-l 


Cluster sampling is also a commonly applied techniques, which has often 
been used in extensive forest surveys, resource assessments and inventories, 
particularly in the tropics. With cluster sampling, the elementary units, on 
which the observations are to be made, are grouped in clusters of pre-assigned 
size. When all elementary units of the cluster are included in the sample we 
have a single phase sampling design. Clusters can be also of unequal sizes. 
The cluster size refers to the number of elementary units that compose the 

Like in double-phase sampling, plots which are grouped in clusters, 
reduce the overall travel distance. However, a cluster sampling design - when 
compared to simple random sampling - is efficient only if the variance within 
clusters is large relative to the variable observed. With cluster sampling 
the variance of estimate is generally larger than that obtained by a simple 
random sampling of the same intensity. This increase in variance is due to 
the correlation between units within clusters. 



The following examples of volume regresasions for selected mangrove 
species are listed in geographical order, starting with Latin America and 
ending with Asia and the Pacific. The list is by no means complete and some 
of the regressions are based on limited measurements and should therefore be 
treated with caution. However, it is hoped that it may be of use for 
preliminary assessments in areas where no volume tables are available. 



Avicennia germinans : Equation: Log(V) = - 9.06038 + 2.39559 x Log(D) 

Correlation coefficient 0.995 

Standard error of estimate 0.149 

F - ratio 4473.063 

Degree of freedom =49 

Laguncularia racemosa: Equation: Log(V) * - 8.72393 + 2.36491 x Log(D) 

Correlation coefficient - 0.989 

Standard error of estimate * 0.197 

F - ratio = 1545.063 

Degree of freedom 28 

Rhizophora mangle : Equation: Log(V) - - 8.92114 + 2.38992 x Log(D) 

Correlation coefficient 0.986 

Standard error of estimate - 0.193 

F - ratio - 962.163 

Degree of freedom 28 

Costa Rica 

Pelliciera rhizophorae: Equation: V * - 0.37714 + 0.03200 x D 

Correlation coefficient 0.994 

Standard error of estimate - 0.038 

F - ratio - 316.146 

Degree of freedom 21 

Rhizophora harrisonii: Equation: V - 0.50857 + 0.04116 x D 

Correlation coefficient - 0.993 

Standard error of estimate 0.058 

F - ratio - 294.946 

Degree of freedom 25 

Note: The volume calculated in Costa Rica is Volume under bark. 
' Source: Chong (1988b and 1989b) 


A. 5. 2 AFRICA 

Sierra Leone 

Rhizopttora. racemosa/harrisonii : V - 0.0001 x D a - S47i , or in tabular form: 


Stem volume 
>7 cm 
(m 1 ) 


Stem volume 
(m s ) 























































































































Source: Layche and Amadou (1 

989) (See aJ 

Lso CAM Stuty 5) 


A. 5.3 ASIA 


The following three volume regressions, shown graphically, are based on 
the inventory of the Sundarbans carried out by ODA in 1984 and described in 
Cast Study 3. The equivalent regressions used in an earlier inventory by 
Forestal in 1960 are also shown. The volume equations and tables giving the 
volume for each 5 cm diameter class can be found in the report by Chaffey et 
al. (1985). 

For each of the species the height classes are as follows: 



H 2 
H < 
H < 


The diamter is dbh (in cm) and the volume is volume under bark (in m 3 ) 
to a 10 cm top diameter. 

28 -| 


35 40 45 50 55 60 

A.5.1: Volume regressions for Keora (Somuntla opetola) 









__ Based on felled tree 
measurements. 1984 


2 Height class 

i I 

6 10 12 14 10 IB 20 22 24 26 28 30 32 34 36 38 40 42 

D.b,h . cm 

Figure A.5.2: Volume refrewioiK for Gewa (Excoecaria agaUocha) 












BtMd on fettod trt 
mtaturemfnts, 1984 

Forettal (1960) 

i i i 

8 10 12 14 18 IB 20 22 24 26 28 30 32 34 36 38 40 

b h., cm 

Figure A.5J: Volume regmstons f or Sundri (Heritor* fomet) 



Rhizophora api cula ta/mucrona ta : 


Stem volume 
>7 cm 
(m 3 ) 


Stem volume 
>7 cm 
(m 3 ) 













































































































Mote: Diameter class 8 cm 
Diameter class 9 cm 

7.5-8.4 cm 
8.5-9.4 cm 

The Philippines 

Volume tables for Rhizophora api cula ta, R. mucronata, Bruguiera 
cylindrica, B. gymiorrhiza, Xylocarpus granatim, Lumnitzera littorea and 
Scyphiphora hydrophyllacea can be found in PCARRD (1991) . 


Abdushahid, M.D. Remote Sensing data fisheries resources and mangrove 

1985 forest of the Bangladesh coastal region. Summary of report. 
ICITP. Universitfe Paul Sabatier, CNRS, Toulouse Prance. 56p 

Adams, M.E. Participatory management of Tanzania's, mangroves. Rural 
1992 Development Forestry Network Paper 13e, p. 1-7. GDI, London 

ADB Non Wood Forest Products. Forestry Master Plan for Bangladesh. 
1992 UNDP/FAO BCD 88/025. Asian Development Bank (TA NO. 1355-BAN) 

Adegbehin, J.O. and Nwaigbo, L.C. Mangrove resources in Nigeria: Use and 
1990 management perspectives. Nature & Resources, Vol. 26, No. 2, 
pp. 13-21. 

Aksornkoae, S. Structure, regeneration and productivity of mangroves in 
1975 Thailand. Ph.D. Thesis, Michigan State Univ. 109 p. 

Aksornkoae, S.; Khemnark, C. and Me 11 ink, W.H.H. Mangrove for chacoal. 
1992a Regional Wood Energy Development Programme in Asia, GCP/RAS/- 
13 I/NET, Field Document No. 30. Bangkok, 41 pp. 

Aksornkoae S., Maxwell, G.S., Havanond, S., Panichsuko, S. Plants in 
I992b mangroves. IUCN Regional Office, Bangkok 

Amadou, C.L. Schoolchildren plant mangroves in Sierra Leone. IDAF 
1989 Newsletter No. 9, pp. 43-45. FAO/DANIDA/NORAD Integrated 

Development of Artisanal Fisheries in West Africa. IDAF, 

Cotonou, Benin. 

Angel G. Priego S, Roberto Vandama C, Leda Menendez C. Evaluacion 
1989 preliminar del impacto ecologico-paisajistico en el area de 
estudio, zona del Guanal, Prov. P. del Rio, Proyecto 
TCP/CUB/8851, "Manejo integrado del ecosistema de manglar 11 , La 
Habana, 21 pp. 

Angell, C.L. The Biology and Culture of Tropical Oysters. ICLARM Studies 

1986 and Reviews 13, 42 p. International Center for Living Aquatic 
Resources Management, Manila, Philippines. 

Anon. British Commonwealth forest terminology, Part 1. Empire 
1953 Forestry Assoc. , London. 

Anon. Fishery Division. Review of the status of coastal aquaculture 
1972 in Malaysia. In Pillay, T.V.R., Ed: Coastal aquaculture in the 
Indo-Pacific Region. London. Fishing News (Books), pp. 48-51. 

Anon. Situacion actual del ecosistema de manglar en Cuba. Academia de 
1986 Ciencias de Cuba, La Havana. 77 pp. 

Anon. Environmental Guidelines for Coastal Protection Measures, 
1988 UNEP/UNDP, Nairobi, 18 pp. 

Arroyo, P.J. Propriedades y usos posibles de las mangles de la region del 
1971 Rio San Juan en la Reserva Forestal de Guarapiche (Estado 

Monagas) . Bol. Inst. Lat-Amer. Invest. Capacit, 33-34: 53-76. 

ASEAN/US CRMP,DGF The Integrated Management Plan for Segara Anakan- 
1992 Cilacap, Central Java, Indonesia. ASEAN/US Coastal Resources 

Management Project Technical Publication Series no. 12. ICLARM, 


A.S.P. Manual of remote sensing. 2nd edition. Colwell et al. Eds. 


Baconguis, S.R. and Mauricio, R.A. Forage and livestock production in the 
1990 mangrove forest. CANOPY International 9, pp. 9-11. 

Bagnouls, F. and Guassen, H. Saison seche et indice xerothermique . 
1953 Documents pour les cartes de productions vegetales (Toulouse) , 
Vol. p. 1-47. 

Baker, I. and Kaeoniam, P. (Eds.) Manual of coastal development planning 
1966 and management for Thailand. Thailand Institute of Scientific 
and Technological Research, Bangkok. 

Baltaxe, R. The application of Landsat data to tropical forest surveys. 
1980 FAO, Rome. 

BARC Proceedings of the seminar on Top Dying of Sundri (Heritiera 
1990 fames) trees. Bangladesh Agricultural Research Council, Dhaka. 

Barua, D.K. and Koch, F.G Field Measurements and Remote Sensing in the 
1986 Estuaries for Engineering Development 10 p. Paper presented at 
the Regional Seminar on the Application of Remote Sensing 
Techniques to Coastal Zone Management and Environmental 
Monitoring. Dhaka-Bangladesh 18-26 November 1986. 

Benessalah, D. Manual on Mapping and Inventory of Mangroves. 
1988 Document prepared for FAO Forestry Department, Rome. 

Berenger, A.C. Reconnaissance des terrains salfes par t616dtection. ICITV. 
1985 CNRS. Universite Paul Sabatier. R6sum6 d'un memoire de fin 
d' etudes. 4lp. 

Betancourt, A. Los manglares en Cuba. Revista Forestal Baracoa V.2(2): 
1950 31-54. 

Birkeland, C. Caribbean and Pacific coastal marine systems: similarities 
1990 and differences. Nature & Resources, Vol 26, no. 2, pp. 3-12. 

Birkenhflger, B. Assessment of Ma-Swar/Ribi River Mangrove Area Wood 
1988 Resource. FAO/FO:DP/SIL/84/003, Field Document No. 16, FAO, 
Freetown, Sierra Leone. 

Bishop, Y.M.M., Fienberg, S.E. and Holland, P.V. Discrete multivariate 



Blasco, F. 

Blasco, F. 

Blasco, F. 


theory and Practice. Cambridge, Mass. MIT press. 

Climatic factors and the biology of mangrove plants. In: 
The mangrove ecosystem: research methods. Snedaker and 
Snedaker, Eds. UNESCO, pp 18-35. 

Taxonomic consideration of the mangrove species. In: Snedaker , # 
S.C.,and Snedaker, J.G., Eds. The mangrove ecosystem: research 
methods. Paris, UNESCO, pp. 81-90. 

Lavenu, F., Chaudhury, M.U. and Kerr, Y. Ressources des Forfits 
de mangrove et d'activitfes humaines dans le Delta du Gange 
(Bangladesh). SPOT image/GDTA Fiche OC 3. 

Amnagement D'Ecosyst&mes Forestiers Fragiles. 

Paper presented at the 10th World Forestry Congress, 17-26 

September 1991, Paris, France 

Boonyobhas, C. Mangrove Forest Management in Thailand. No. R.161. pp. 
1975 1-42. Forest Management Division, Royal Forest Dept., Bangkok. 

Borel, D. Digital analysis of high resolution satellite simulation data 
1983 on mangrove, in: ESCAP/UNESCO/NRCT Regional Remote Sensing 

Training Course on Mangrove Ecosystem. Bangkok, Nov. 28 - 

Dec. 16, 1983. pp. 97-105. 


Borel, D. The Design of Geo-Referenced Data Processing Systems in 

1984 Mult i- level Remote Sensing for Forestry Applications pp. 
233-245. ESCAP/NRMC Proceedings of Regional Conference on 
mult i- level Remote Sensing for Forestry Application. 8-12 
October 1984. Quezon City, Metro Manila, Philippines. 

Borel, D. Satellite based remote sensing systems : The on going programs, 

1985 Landsat-4/5 and SPOT. In: Remote sensing in vegetation 
studies: Report of the training course on remote sensing 
techniques applied to vegetation studies. Biotrop, Bogor, 
Indonesia, pp: 77-87. 

D., and Chaudhury, M.U Monitoring of Mangrove Areas through High 
Resolution Remote Sensing Techniques; The Spot Simulation 
Campaign over Bangladesh, pp. 106-116. 


Bradbear, N. Beekeeping in rural development. 
1990 Association, Cardiff, U.K. 

International Bee Research 

Brasnett, N 

Broom, M.J. 

Bunt , J . S . , 

Burchett, M 

Burchett, M 

Caioli, A.B 

Carlos P. 

Carson, G.L 

Carter, M.R 

Castellu, J 

Chacko, V.J 

Chaffey, D. 

.V. Planned management of forest. Allen & Unwin, London. 
238 pp. 

The Biology and Culture of Marine Bivalve Molluscs of the 
Genus Anadara. ICLARM Studies and Reviews 12. International 
Center for Living Aquatic Resources Management, Manila, 

Boto, K.G. and Boto, G. A survey method for estimating 
potential levels of mangrove forest primary production. Marine 
Biology. 52: 123-128. 

.D. Mangrove management policies, In: Mangrove Ecosystems of 
Asia and the Pacific - status, exploitation and management. 
Eds: Field, C.D. and Dartnall, A.J., Proc. of the Research for 
Development Seminar, Aust.Inst. Marine Science, Townsville, 
Australia, pp. 305-320. 

.D., Field, C.D. and Pulkownik, A. Salinity, growth and root 
respiration in the grey mangrove, Avicennia marina. Physiol. 
Plant. 60: 113-118 

Philippine Mangrove Conservation. The Application of Multi- 
level remote sensing for natural resources decision making. 
Seminar Proceedings Beijing 14-24 April 1985. pp. 175-183. 

Informe sobre los aspectos pesqueros, Proyecto TCP/CUB/8851, 
"Manejo integrado del ecosistema de manglar", La Habana, 17 pp. 

. The "Spot" Method of Enumeration Survey in Matang Mangroves. 
The Malayan Forester. Vol. 16, no. 4. pp. 206-212. 

. , Burns, L.A., Cavinder, T.R., Dugger, K.R., Fore, P.L., Hicks, 
D.B., Revells, H.L. and Schmidt, T.W. Ecosystem analysis of 
the Big Cypress Swamp and estuaries. U.S.E.P.A. Region IV, 
South Florida Ecology Study. 

.P. Microwave and vegetation. In: Remote Sensing in vegetation 
studies. Report of training course on Remote Sensing 
techniques applied to vegetation studies. Biotrop, Bogor, 
Indonesia. Nov. 4 - Dec. 13, 1985. pp. 65-76. 

. A manual on sampling techniques for forest surveys. Survey 
of India Off. (P.Z.O.), Dehra Dun. 172p. 

R., Miller, F. R. and Sandom, J. H. A forest inventory of the 
Sundarbans, Bangladesh, Main Report, ODA. Project Report 140. 


Chan, H.T. The Growth of Mangrove Species (unpublished) . 

Chan, H.T., Nasir Husin and Chong, P. Is there a need to eradicate 
1987 Acrostichum speciosim prior to planting of Rhizophora mucronata 
in logged-over mangrove forest area?, PRIM Occasional Paper, 
Forest Research Institute, Kepong, Malaysia, No. 1/87, 7 pp. 

Chan, H.T.; Razani U. and Putz, F.E. Growth and natural regeneration of a 
1982 stand of Rhizophora apiculata trees at the Matang mangroves 

following thinning. In: Symposium on Mangrove Forest Ecosystem 

Productivity. Bogor, Indonesia, 10 pp. 

Chan, H.T. and Sallleh, M.N. Traditional uses of the mangrove ecosystem in 
1987 Malaysia. UNDP/UNESCO Regional Mangroves Project RAS/86/120. 
Mangroves Occasional Papers no.l 

Chapman, V.J. 

Mangrove phytosociology. Trop. Ecol. 11(1): 1-19. 

Chapman, V.J. Mangrove biography. In: Walsh, G.E. ,Snedaker, S.C. and Teas, 
1974 H.J., Eds. Proceed, of the intl. symp. on biol. and mgmt. of 
mangroves. Florida. Univ. of Florida. 6 pp. Iv. 

Chapman, V.J. 

Mangrove vegetation, Vaduz., J, Cramer, 197 pp. 

Chapman, V.J. Botanical surveys in mangrove communities. In : The mangrove 
1984 ecosystem: research methods pp. 53-80. Snedaker and Snedaker, 

Chaudhury, M.U. Remote Sensing Application to the Study of Mangrove 

1983 Ecosystem pp. 56-66. ESCAP/UNESCO/NRCT Regional Remote Sensing 
Training Course of Mangrove Ecosystem Bangkok, Nov. 28 -Dec. 16 

Chaudhury, M.U. Multi-level Remote Sensing for Forest Vegetation Mapping. 

1984 pp. 247-252. ESCAP/NRMC Proceedings of Regional Conference on 
multi-level Remote Sensing for Forestry Application. 8-12 
October 1984. Quezon City, Metro Manila, Philippines. 

Chaudhury, M.U. Lands^t series: Technical properties and application to 

1985 vegetation studies. In : Remote Sensing in Vegetation Studies. 
Report of training course on Remote Sensing techniques applied 
to vegetation studies. Biotrop, Bogor, Indonesia. 
Nov.4-Dec.13, 1985. pp. 65-76. 

Chaudhury, M.U. Application of Satellite Remote Sensing Technology to 

1986 Mangrove Vegetation Analysis in Bangladesh. Paper presented at 
the Regional Seminar on the Application of Remote Sensing 
Techniques to Coastal Zone Management and Environmental 
Monitoring. Dhaka-Bangladesh 18-26 November 1986. '15p. 

Chaves, E.S. and Fonseca, W.G. Plan de mane jo para mangle en Coronado de 

1986 Osa, Puntarenas. Heredia. 50 pp. 

Chevrou, R.B. Precision des mesures de superficie estimSe par grille de 
1976 points ou intersections de parallSles. Ann. Sci. For. 
33(4) :257-269. 

Chevrou, R.B. The precision of dot grid estimates: A theoretical approach. 
1979 Resources Inventory notes, BLM 20:3-6. 

Chong, P.W. Proposed management and integrated utilization of mangrove 

1987 resource in Sierra Leone, Freetown, Project UNDP/FAO: 
SIL/84/003. Field Document No. 6, 123 pp. 


Chong, P.W. Forest management plan for Playa Garza pilot area: Terraba- 
1988a Sierpe Mangrove Reserve, San Jose, Project FAO/TCP/COS/6652. 
Field Document No. 3. 76 p. [also in Spanish] 

Chong, P.W. Proposed Integrated Forest Management Planning and Utilization 
1988b of mangrove resources in the Terraba- Sierpe Reserve, Costa 

Rica. Proj. FAO/TCP/COS/6652 Technical Report 2. 150 pp. [also 

in Spanish] 

Chong, P.W. A forest management plan for Dat Mui Forest Enterprise, Hanoi, 
1989a Project UNDP/FAO: VIE/82/002, Field Document No. 9., 44 pp. 

Chong, P.W. Mane jo integrado de ecosistemas de manglares en la reserva 
1989b forestal de Guanal - Cuba. Programa de Cooperacion Tecnica; 

Documento de Campo 1. (FO:TCP/CUB/8851) , 138 p. Roma, [also in 

English] . 

Chong, P.W. Mangrove Forest Management Guidelines. Paper prepared for the 
1993 FAO Forestry Department, FAO, Rome. 

Chong, V.C.,and Sasekumar, A. Food and feeding habits of the white prawn 

1981 Penaeus merguiensis. Marine Ecol. Progress Series, 5: 185-191. 

Christensen, B. Biomass and primary productivity of Rhizophora apiculata 
1978 Bl. mangrove in Southern Thailand. Aquatic Botany, 4: 43-52. 

Christensen, B. Management and utilization of mangroves in Asia and the 

1982 Pacific, Rome, FAO Environment Paper No. 3. 160 pp. 

Christensen, B. Mangroves, what are they worth? Unasylva vol. 35 no. 139. 

1983 pp. 2-15. 

Christensen, B. and Snedaker, S.C. Integrated Development of the 

1984 Sundarbans: Ecological aspects of the sundarbans. FAO: Field 
Document No. 3, FO:TCP/BGD/2309(Mf) , Rome. 

Christensen, B. and Wium- Andersen, S. Seasonal growth of the mangrove 

1984 trees in Southern Thailand. I. The phenology of Rhizophora 
apiculata Bl. Aquatic Botany, 3: 281-286. 

Chou, L.M., Chua, T.E., Khoo, H.W., Lim, P.E., Paw, J.N, Silvestre, G.T., 
1991 Valencia, M.J., White, A.T. and Wong, P.K (Eds.) Towards an 
Integrated Management of Tropical Coastal Resources. 
Proceedings of the ASEAN /US technical Workshop on Integrated 
Coastal Zone Management, 28-31 Oct. 1988, Singapore. ASEAN/US 
CRMP Conference Proceedings no. 4. ICLARM, Manila. 

Chukhamdee, J. Application of Aerial Photograph for Estimating Mangrove 
1986 Forest Production in Ranong Province. MSc. Thesis. 

Cintron, G. Mangrove Area Management in Panama - Final Report. Puerto 

1985 Rico, P.O. 40-319R-5-00194, USDA. OICD. 44 pp. 

Cintron, G., Lugo, A.E., Pool, D.J. and Morris, D. Mangroves of arid 
1978 environments in Puerto Rico and adjacent islands. Biotropica 
10: 110-121. 

Cintrfin, G. and Novelli, Y.S Methods for studying mangrove structure. 
1984 In: The mangrove ecosystem: research methods, pp. 91-112. 
Snedaker and Snedaker, Eds. UNESCO 

Cintron, G.Y., and Schaeffer-Novelli, Y. Introduccion a la ecologia del 
1983a manglar. Montevideo, UNESCO, 99 pp. 


Cintron, G.Y. and Schaeffer-Novelli, Y v Mangrove forests: ecology and 
1983b response to natural and induced stressors. In: Ogden, J.C. and 
Gladfelter, E.H., Eds. Coral reefs,, seagrass beds and mangrove: 
their interaction in the coastal zones of the Caribbean. UNESCO 
Reports in Marine Science No. 23. UNESCO, pp. 87-104. 

Clarke, L.D.,and Harmon, N.J. The mangrove swamp and salt marsh 
1970 communities of the Sydney District: 111. plant growth in 

relation to salinity and waterlogging. Jour, of Ecol. 58: 


Clifford, H.C., Semi -intensive shrimp farming. In: Chamberlain, G.W., 
1985 Haby,M.G., and Miget, R.J. Eds. Shrimp farming manual. Corpus 
Christ i, Texas, Texas Agric. Exten. Services, IV: 15-42. 

Clough, B.F. Growth and salt balance of the mangroves Avicennia marina 

1984 Forsk.) Vierh. and Rhizophora stylosa Griff, in relation to 
salinity. Aust. Jour. Plant, Physiol. 11: 419-430. 

Clude, J.B. The potential for mollusc farming in Costa Rica. A report to 

1981 the aquaculture development and coordination programme of FAO. 
Seattle, Glude Aquaculture Consultants Inc. 29 pp. 

Cochran, W.G. Sampling techniques. 3rd edition. John Wiley and Sons, 

1977 New York. 428p. 

Cohen, J. A coefficient of agreement for nominal scales. Educational 
1960 and Psychological Measurement. 20(l):37-46. 

Cohen, J. Weighted Kappa: Nominal scale agreement with provision for 

1968 scale disagreement or partial credit. Psy. Bui. 70(4): 213-220. 

Congalton, R.G., Oderwald, R.G and Mead, R.A Accuracy of remotely sensed 

1982 data: Sampling analysis procedures. Remote sensing research 
report. 82-1. Agristars Rep. RR-U2- 04257. Coop. Agreement 
13-1134. VPI and State Univ. 83p. 

Connor, D.J. Growth of the grey mangrove Avicennia marina in nutrient 

1969 culture. Biotropica 1: 36-40. 

Cook, H.L. and Rabanal, H.R. (eds.) Manual on pond culture of penaeid 

1978 shrimp. ASEAN National Coordinating Agency of the Philippines., 
Min. of Foreign Affairs, Manila. 132 pp. 

Cundell, A.M., Brown, M.S., Stanford, R., and Mitchell, R. Microbial 

1979 degradation of Rhizophora mangle leaves immersed in the sea. 
Estuarine and Coastal Marine Science, 9, 281-286. 

Dadant and Sons (eds), The Hive and the Honey Bee, Hamilton, Illinois: 
1982 Dadant and Sons Inc. 

Das, S. and Siddiqi, N.A. The Mangroves and Mangrove Forests of 

1985 Bangladesh. Mangrove Silviculture Division Bulletin No. 2. 
Bangladesh Forest Research Institute and UNDP/FAO Project 
BGD/79/017. Chittagong, Bangladesh. 

Davis, J.H.Jr. The ecology and geological role of mangroves in Florida. 
1940 Pap. Tortugas Lab., 32: 303-412. 

Dawkins, H.C. The management of natural tropical high- forest, with special 
1958 reference to Uganda. Imperial For. Inst. Paper No. 34,155 pp. 

De Haan, J,D. Het een en auder over de Tijlatjapsche vloedbosschen. 
1931 Tectona, Vol. 24, pp. 39-76. 


Dinh-Hiep, Application of Multi-Level Remote Sensing to Forest Inventory 
1984 and Planning in Viet -Nam. ESCAP/NRMC Proceedings of Regional 
Conference on mult i- level Remote Sensing for Forestry 
Application. 8-12 October 1984. Quezon City, Metro Manila, 

Doat, J. 

Le pouvoir calorifique des bois tropicaux. Revue B.F.T. No. 172. 

Downton, W.J.S. Growth and osmotic relations of the mangrove Avicennia 
1982 marina as influenced by salinity. Aust. Jour. Plant Physiol. 

Dunn, I.G. Notes on mass fish death following drought in Malaya. Malay. 
1975 Agric. Jour. 45: 204-211. 

Durant, C.C.L. The growth of mangrove species in Malaya. The Malayan 


Forester. Vol. 10. pp. 3-15. 

Dwivedi, S.N., Parulekar, A.H., Goswami, S.C. and Untawale, A.G., Ecology 
1974 of mangrove swamp on the Mandovi Estuary, Goa, India. In: 

Walsh, G.E., Snedaker, S.C. and Teas, H.J., Eds. Proceed, of 
the intl. symp. on biol. and mgmt. of mangroves. Florida, Univ. 
of Florida, pp. 115-125. 

Edwards, R.R.C. Field experiments on growth and mortality of Penaeus 

1977 vannamei in a Mexican coastal lagoon complex. Estuar. Coast. 
Mar. Sci., 5: 107-121. 

Edwards, R.R.C. The fishery and fisheries biology of penaeid shrimp on 

1978 Pacific coast of Mexico. Oceanogr. Mar. Biol. 16: 145-180. 

Ellis, E.M. Moluscos de Nicaragua y Costa Rica. Programa regular de 
1968 Desarrollo Pesquero de Centroamerica, Informes. 8 pp. 

Erftemeijer, P., van Balen, B., Djuharsa, E. The importance of Segara 

1988 Anakan for nature conservation, with special reference to its 
avifauna. Asian Wetland Bureau/INTERWADER - PHPA. Bogor 

Ero, I.I and J.E. Abu, Consideration of Mangrove Forest management in 1991 
Nigeria. Voluntary paper for the Senior Foresters Conference, 
ITTO', Yokohama, Japan, 23-26 July, 1991. 

ESCAP Coastal Environmental Plan for Bangladesh. Volume two: Final 
1987 Report. Economic and Social Commission for Asia and the 
Pacific. Bangkok 

Esteban C. Menendez P. Inventario Forestal: Resumen de methodos y 

1989 tecnicas apropiadas para el ecosistema manglar. Experiencias 
practicas en el area el Guanal, EFI C. Sur, Prov. P. del Rio. 
Proyecto TCP/CUB/8851 "Manejo integrado del ecosistema de 
manglar 11 , La Habana, 38 p. 

Estribi, C., Palacios, T. and Gonzalez, J. Informe de la situacion actual 
1983 de la produccion de casacara de mangle en la provincia de 
Chiriqui. RENARE/MIDA. Panama. 14 p. 

Eugenio Poussin Molinet. Explotacion forestal del manglar: Introduccion de 
1989 technologia mechanizada. Proyecto TCP/CUB/8851. H Manejo 
integrado del ecosistema de manglar ", La Habana, 28 p. 

Evink, G.L. Macrobenthos comparisons in mangrove estuaries. In: Walsh, 
1974 G.E., Snedaker, S.C., and Teas, H.J., Eds. Proceed, of the 

intl. symp. on biol. and mgmt. of mangroves. Florida. Univ. of 

Florida, 257 pp. Iv, 


FAO. Tropical silviculture. Vol. l. 

FAO A framework for land evaluation, FAO Soils Bull. 32, Rome ,72 pp. 

FAO Management and utilization of mangroves in Asia and the 
1982 Pacific; based on the work of Christensen, B. Environ. Paper 3, 
160 pp. 

FAO Land evaluation for forestry, FAO Forestry Paper 48, Rome, 
1984a 123 pp. 

FAO. Mangrove Forests in Asia- Pacific Region. Special Study on 
1984b Forest Management, Afforestation and Utilization of forest 
resources in the developing regions. GCP/RAS/106/JPN. 38p. 

FAO Apoyo a la ejecucion de los programas forestales prioritarios: 
1985a Administracion y organizacion forestal. based on the work of 

Reyes Cerrato, R. FO.DP/COS/79/001. Technical Report No. 7, 

Rome, 35 p. 

FAO. Mangrove Management in Thailand, Malaysia and Indonesia. 
1985 FAO Environment Paper 4. FAO, Rome. 

FAO Mane jo integral de un area de manglar: Costa Rica, Reserva 
1988 Forestal Terraba-Sierpe. Informacion basica. based on the work 
of Martin, I.N. TCP/COS/6652:FAO/DGF. Technical Report, 123 p. 

FAO The management of crocodiles in captivity. FAO Conservation 

1990 Guide 22. FAO, Rome 

FAO Mane jo Integrado de Ecosistema de Manglares. Cuba. 

1991 Final Report, FAO/FO:TCP/CUB/8851, Rome. 

FAO Integrated management of coastal zones. FAO Fisheries 
I992a Technical Paper no. 327. FAO, Rome 

FAO Mangrove for Charcoal. A Vanishing Woodfuel Resource System. 
1992b (The Case of Yeesarn, Upper Gulf of Thailand) 

FAO/GCP/RAS/131NET Field Document No. 30, Bangkok. 

FAO/MINAGRI, Informe de CUBA: Seminario sobre "Mane jo integral de 
1984 ecosistemas de manglares", Manzanillo, 32 p. 

FAO/UNDP Mise en valeur de la basse et moyenne Casamance. Reublique du 
1975 Senegal, Inventaire Forest ier, Fascicule 1: Description de la 
region inventoiree, Methode et realisation de I 1 inventaire, 
Resultats d' ensemble et commentaires. DP/SEN/71/522-l/FO. 
FAO, Rome. 

FAO/UNEP tropical Forest Resources Assessment Project: Forest Resources 
1981a of Tropical America. UN 32/6.1301-78-04. Technical Report 1, 
FAO, Rome 

FAO/UNEP Tropical Forest Resources Assessment Project: Forest Resources 
1981b of Africa. UN 32/6.1301-78-04. Technical Report 2. FAO, 

FAO/UNEP Tropical Forest Resources Assessment Project: Forest Resources 
1981C of Tropical Asia. UN 32/6.1301-78-04. Technical Report 3. 
FAO, Rome. 


Fell, J.W., Cefalu, R.C., Master, I.M. and Tallman, A.S. Microbial 
1975 activities in the mangrove (Rhizophora mangle) leaf detrital 
system. 'Proceed, of the intl. symp. on biol. and mgmt. of 
mangroves', In: G. Walsh, Snedaker, S,C,, and Walsh, G.E., 
Eds., Florida, Gainesville, Univ. of Florida, pp. 661-679. 

Fitzpatrick-Lins, K. Comparison of sampling procedures and data analysis 
1981 for a land use and land cover map. Photo. Eng. Rem. Sen. 
47(3) : 343-351. 

Fleiss, J.L., Cohen, J. and Everitt, B.S. Large sample standard errors of 
1969 Kappa and Weighted Kappa. Psy. Bui. 72 (5) ,-323-327. 

Fox III, L. "Image Smoothing" An Aid to Resource Mapping in the Tropics. 


pp. 87-92. In: ESCAP/UNESCO/NRCT Regional Remote Sensing 
Training Course on mangrove ecosystem. Bangkok, Nov. 28- 
Dec.16, 1983. pp. 97-105. 

Frayer, W.E. Multi-Level Sampling Designs for Resource Inventories. Rocky 


Freese, F. 

Frisk, T. 

Frodin, D. 

Mountain Forest and Range Experiment Station 
pp. 111. 

USDA For. Ser. 

Elementary forest sampling. Agr. handbook, U.S. Dept. Agri. 
For. Ser. 91 p. 

Some observations on harvesting mangrove forests in Peninsular 
Malaysia. FO.-MISC/84/18 . FAO, Rome 

The mangrove ecosystem in Papua New Guinea, In: Field, C.D., 
Dartnell, A.J., eds. Mangrove ecosystems of Asia and the 
Pacific, Townsville, Australia. 53-63. 

Gelnet, R.H., Dellwig, L.F and Bare, J.E Increased visibility from the 
1978 invisible. A comparison of Radar and Landsat in tropical 
environment. In: Proceedings of the 12th International 
Symposium on Remote Sensing of Environment, pp. 2205-2216. 

Gill, L.S. Report on Mangrove Management of the Ayeyarwady Forest. 
1992 Feasibility Study on Mangrove Reforestation, FAO/FO:MYA/90/003 
Field Document No. 3 f Yangon. 

Gocke, K., Vitola, M. and Rojas, G. Oxygen consumption in a mangrove swamp 
1981 on the Pacific coast of Costa Rica. Revista de Biologia 
Tropical 29(1) : 143-155. 

Groulez, J. L' utilisation de 1' avion dans les reconnaissances et 
1954 prospect ions forest iSres au Gabon. Revue Bois et Forfets des 
Tropiques no. 35, pp. 25-32. 

Grubb, P.J. The maintenance of species richness in plant communities, the 
1977 importance of the regeneration niche. Biological Reviews, 52: 

Haan, J.H. de. Het een en ander over de Tjilatjapische vloedbosschen, 
1931 Tectona 24: 39-75. 

Hagberg, A., and Kalb, C. Marine shelled mollusks of commercial 
1968 importance in Central America, Boletin tecnico. FAO. El 
Salvador. 32p. 

Hamilton, L.S. and Snedaker, S.C. Eds. Handbook for Mangrove Area 
1984 Management. lUCN/Unesco/UNEP. East-West Center, Honolulu, 
Hawaii. 123 pp. 

Hansen, M.H., Hurwitz, W.N. and Madow, W.G. Sample survey methods and 
1953 Methods. Vol. I. J. Wiley and Sons, New York. 638 p. 


Karon! H.A.H. A working plan for the second 30-year rotation of the 
1981 Matang Mangrove Forest Reserve, Perak. The first 10-year 

period 1980-1989. Perak State Forestry Dept., Perak, Malaysia. 

109 p. 

Hay, A.M. Sampling designs to test land use map accuracy. Photo. Eng. 
1979 Rem. Sen. 45(4) :529-533. 

Hesse, P.R. Some differences between the soils of Jlhizophora and Avicennia 
1961 mangrove swamps in Sierra Leone. Plant and Soil, 14, 249. 

Heald, E.J. The production of organic detritus in a south Florida estuary. 


Sea Grant Techn, Bull. No. 6., Univ. of Miami, Sea Grant Program 
(Living Resources), Miami, Florida HOp 

Heald, E.J. and Odum, W.E. The contribution of mangrove swamps to Florida 

1970 fisheries. Proceed, of the Gulf and Caribbean Fisheries Inst., 
22, 130-135. 

HKUST Programme and Abstracts, Asia-Pacific Symposium on Mangrove 
1993 Ecosystems, Hong Kong 1-3 Sept. 1993. Hong Kong University of 
Science and Technology, Hong Kong. 

Hoang, V.G. A study on the use of computer processing of Landsat MSS Data 
1986 for Coastal Zone Mapping, lip. Paper presented at the 
Regional Seminar on the Application of Remote Sensing 
Techniques to Coastal Zone Management and Environmental 
Monitoring. Dhaka-Bangladesh 18-26 November 1986. 

Hoffman, W.E. and Dawes, C.J. Photosynthetic rates and primary production 
1980 by two Florida benthic red algal species from a salt marsh and 
a mangrove community. Bull, of Mar. Science 30: 358-364. 

Holdrige, L.R., Grenke, W.C., Hatheway, W.H., Liang, T. and Tosi, J.A. Jr. 

1971 Forest Environments in Tropical Life Zones: A pilot study. 
Pergamon Press. 

Hord, R.M. and Brooner, W. Land use map accuracy criteria. Photo. Eng. 
1976 Rem. Sen. 42 (5):671-677. 

Hossain, D. Remote Sensing Technology applied to Ecological Studies of 
1985 Mangrove Ecosystems in coastal regions of Bangladesh. ICITV. 
CNRS. University Paul Sabatier. Summary report 37p. 

Hughes, D.G. Aquaculture: Central America. In: Proceed. Shrimp world 

1985 market conference. Acapulco, Mexico, Nov. 30 - Dec. 2 1984. New 
Orleans, Louisiana, Shrimp World Incorporated, pp. 162-166. 

Hush, B., Miller, C.I and Beers, T.W. Forest Mensuration, 2nd edition, 

1972 Ronald Press Company, New- York. 410 p. 

Hussain, M.Z. Restoration and Expansion of the mangrove belt in Guyana. 
1990 TCP/GUY/8953. Technical Report No.l. FAO, Rome. 

Hussain, M.Z. Multiple -use management of the Sundarbans Forest in 
1992 Bangladesh. Case Study prepared for the FAO Forestry 
Department, FAO, Rome. 

Imhoff , M.L. and Vermillion, C.H. The Role of Space Borne Imaging 

1986 Radars in Environmental Monitoring: Some Shuttle Imaging Radar 
Results in Asia. NASA/Goddart Space Flight Center Greenbelt, 
MD. 19 p. 


Insull, A.D.; Barg, U.C and Martosubroto Coastal fisheries and coastal 
1993 aquaculture within integrated coastal area management in East 
Africa. Paper presented at the Workshop/Conference on 
integrated coast el zone management in East Africa, Arusha, 
Tanzania 21-23 April, 1993. Fao, Rome. 

Ishaq-Mirza, M., Hasan, M.Z., Akhtar, S., Ali, J. and Sanjrani, M.A. Remote 
1986 Sensing Survey of Mangrove Forest Along the Coast of 

Baluchistan. I2p. Paper presented at the Regional Seminar on 
the Application of Remote Sensing Techniques to Coastal Zone 
Management and Environmental Monitoring. Dhaka-Bangladesh 
18-26 November 1986. 

ISME/ITTO Proceedings of a workshop on Conservation and Sustainable 
1993a Utilization of Mangrove Forests in Latin America and Africa 

Regions. ITTO/ISME Project PD114/90(F) Niteroi, Brazil 28-30 
May, 1993. Mangrove Ecosystems Proceedings no. 1. ITTO SRS 
No. 12. ISME, Japan 

ISME/ITTO Proceedings of a workshop on Conservation and Sustainable 
I993b Utilization of Mangrove Forests in Latin America and Africa 
Regions. ITTO/ISME Project PD114/90(F) Dakar, Senegal 20-22 
January, 1993. Mangrove Ecosystems Proceedings no. 2. ITTO SRS 
No. 13. ISME, Japan 

Jano, A. Large scale photography for forest inventory problems and 
1979 limitations. Paper presented in Workshop on practical 

applications of remote sensing to timber inventory. Edmonton, 

Alberta. Sept. 26-28, 1979. 

Jimenez, J.A. A hypothesis to explain the reduced distribution of the 

1984 mangrove Pelliciera rhizophorae Tr. & PI., Biotropica 16(4): 

Jimenez, J.A. and Soto, R. Pat rones regionales en la estructura y 

1985 composicion floristic de los manglares de Costa Rica. Rev. 
Biol. Trop. 33(1) . 

Johns tone, I.M. and Frodin, D.G. Mangroves of the Papuan subregion. In: 

1982 Biogeography and ecology of New Guinea. Monographiae biologicae 
42. Gressit, J.L. Ed. The Hague, pp. 513-528. 

Jothy, A. A., Capture fishery and the mangrove ecosystem: In: Ong. J.E., 

1983 and Gong W.K. Eds., Productivity of the mangrove ecosystem: 
Management implications. UNESCO/UNDP (RAS/79/002/G/01/13) , 
Penang, pp. 129-141 

Kanapathy, K. The reclamation and improvements of acid sulphate soils for 
1975 agriculture. Malay. Agric. Jour. 50: 264-290. 

Keen, A. Sea shells of Tropical West America, Stanford Univ. Press. 
1971 1064 pp. 

Khan, F.A., Choudhury, A.M and Islam, J. A comprehensive report on the 

1986 training exercise on forest survey system and field work 
procedure. Bangladesh Space Research and Remote Sensing 
Organization. (SPARRSO) pp. 1-25. 

Khan, F.A., Choudhury, A.M. and Islam, M.J. Timber Volume Inventory in the 
1990 Sundarbans Using Aerial Photography and other Remote Sensing 

Techniques. UNDP/UNESCO Regional mangroves project RAS/86/120. 

Mangrove Ecosystems Occasional Papers No. 9. 

Kjerfve, B. Hydrographic considerations in estuarine outwelling studies: 

1984 An example and definitions. In: Ong. J.E., and Gong W.K. Eds., 
Productivity of the mangrove ecosystem: Mgmt. implications. 
UNESCO/UNDP (RAS/79/002/G/01/13) , Penang, pp. 37-47. 


Kjerfve, B, Manual for investigation of hydrglogical processes in mangrove 

1990 ecosystems. UNDP/UNESCO Regional mangroves project. RAS/86/120 

Klankamsorn, B. and Thongchai, C. Principles of Photointerpretation. pp. 

1983 34-40. ESCAP/DNESCO/NRCT Regional Remote Sensing Training 
Course on Mangrove Ecosystem. Bangkok, Nov.28-Dec.16 1983. 

Kunstadter, P., Bird, E.C.F. and Sabhasri, S. (Eds.) Man in the Mangroves 
1987 The Socio-economic Situation of Humen Settlements in Mangrove 
Forests. NRTS-29/UNUP-607 

Lahiri, A.K. Aerial seeding in mangrove swamps. The Indian Forester 

1991 Vol. 117, no. 3 pp. 159-161. 

Lai, P.N. Ecological Economic Analysis of Mangrove Conservation. A Case 
1990 Study from Fiji. UNDP/UNESCO Regional Mangroves Project 
RAS/86/120. Mangrove Occasional Papaers NO. 6. 

Landon, F.H. Mangrove Volume Tables. The Malayan Forester. Vol. 11, no. 3. 
1948 pp. 117-120. 

Landon, F.H. Malayan tropical rain forest. Malay Forester, 18, 30-38. 

Lanly, J.P. Manual of Forest Inventory with special references to tropical 
1973 tropical forests. FAO, Rome. 200 p. 

Lantieri, D. Evaluation de donnfees SPOT simulfees pour 1' Etude de 
1986 diffferents thSmes lies aux ressources renouvelables. FAO, RSC 
Series 37. 85 p. 

Lebigre, J.M. Les mangroves des rias du littoral Gabonais: Essai de 
1982 cartographic typologique. Revue Bois et ForSts des Tropiques. 
No. 100, ler trimestre, pp. 3-28. 

Leda Menendez C, Roberto Vandama C., Ecologia de los manglares, Proyecto 
1989 TCP/CUB/8851. tf Manejo integrado del ecosistema de manglar w , La 
Habana, 34 p 

Leh, C.N.U. and Sasekumar, A. Feeding ecology of prawns in shallow waters 

1984 adjoining mangrove shores. In: Soepadmo, E., Rao, A.N. and 
Macintosh, D.J., Eds. Proceed, of the UNESCO/Asian symp. on 
mangrove environment: Research and management. Univ. Malaya, 
pp. 331-353. 

Letourneau, L.R. and Dixon, R.G. Panama mangrove management and 

1982 harvesting, FAO, RLA/77/019, Working document No. 82/44, 22 pp. 

Lillesand, T.M. and Kiefer, R.W. Remote sensing and image interpretation. 
1979 J.Wiley & Sons. 612p. 

Little, E.L. Common Fuelwood Crops. A Handbook for Their Identification. 
1978 Communi-Tech Associates, Morgantown, West Virginia. 

LoubersaCi L. Remote Sensing of temperate and tropical intertidal zones 

1983 using SPOT simulated data. In: The proceedings of the 
Seventeenth International Symposium on Remote Sensing of 
Environment. Ann Arbor, Michigan, May 9-13, 1983. 

pp. 1363-1371. 

Loetsch, F. and Haller, E. Forest Inventory Vol. 1, BLV 
1973 Verlagsgesellschaft, Mflnchen. 436 p. 

L0yche, M. Mangrove Management and Utilization. Report on a Study Tour 
1989 to Malaysia and Indonesia. Field Document 3, FAO/UNDP Project 
SIL/88/008. Freetown, Sierra Leone. 


Loyche, M. and Amadou, C.L. A Tentative Volume Table for the Mangroves of 
1989 Sierra Leone. Field Document 1, PAO/UNDP Project SIL/88/008. 
Freetown, Sierra Leone. 

L0yche, M. Protection of sea dykes through planting of trees. Appraisal 




of the forestry component of WFP project Vietnam 4617 
Rehabilitation and Upgrading of Sea Dykes. WFP/FAO, Rome 

Mangrove ecosystems: Successional or steady state?. Biotropica 
12: 65-72. 

Lugo, A.E. and Cintron, G. The mangrove forests of Puerto Rico and their 
1975 management. In: G. Walsh, Snedaker, S, and Teas, H. Eds. 

Proceed, intl. symp. on biol. and mgmt. of mangroves. IFAS, 
Florida. Gainesville. Univ. of Florida, pp. 825-846. 

Lugo, A.E., Evink, G., Brinson, M.M.,Broce, A. and Snedaker, S.C. Diurnal 
1975 rates of photosynthesis, respiration and transpiration in 
mangrove forests in south Florida. In: Tropical ecological 
systems. Golley, F. and Medina, G. Eds. Springer-Verlag, N.Y. 
pp. 335-350. 

Lugo, A.E., Sell, M and Snedaker, S.C. Mangrove ecosystem analysis. In: 

1973 U.S.A. University of Florida. The role of mangrove ecosystems 
in the maintenance of environmental quality and high 
productivity of desirable fisheries. Gainesville, Florida. 

Lugo, A.E. and Snedaker, S.C. The ecology of mangroves. Ann. Rev. of 

1974 Ecol. and System. 1974, 5: 39-64. 

Lugo, A.E. and Snedaker, S.C. Properties of a mangrove forest in southern 

1975 Florida. In: G. Walsh, Snedaker, S, and Teas, H. Eds. Proceed. 
intl. symp. on biol. and mgmt. of mangroves. IFAS, Florida. 
Gainesville. Univ. of Florida, pp. 170-212. 

Luna Lugo, A. Mane jo de manglares en Venezuela. Institute Forestal 

1976 Latino- Americano. De investigacion y capacitacion. Merida- 
Venezuela. Boletin NO. 50. p. 41. 

Macnae, W. Mangroves in Eastern and Southern Australia. Aust. Jour. Bot. 
1966 15: 67-104. 

Macnae, W. A general account of the fauna and flora of swamps and forests 
1968 in the Indo-West-Pacif ic Region. Adv. Mar. Biol. 6: 73-270, 

Macnae, W. Mangrove forests and fisheries. FAO/UNDP Indian Ocean 
1974 Programme, Indian Ocean Fishery Commission IOFC/DEV/74/34, 
Rome, 35 pp. p. 

Marius, C. Mangroves du S6n6gal et de la Gambie. Ecologie-Pfedologie 
1986 GSochimie. Mise en valeur et amfinagement. pp 231-247. 

Martosubroto, P. and Naamin, N. Relationship between tidal forests 

1977 (mangroves) and commercial shrimp production in Indonesia. Mar. 
Res. in Indonesia 18: 81.-86. 

Mattes, H. and Kapetsky, J.M. World wide compendium of mangrove-associated 
1988 aquatic species of economic importance. FAO Fish. Circ. No 814 
FAO, Rome 

Meisner, D.E. and Lindstrom, O.M. Design and operation of a Color 
1985 Infrared Aerial Video System, Photogrammetric Engineering and 
Remote Sensing. Vol. 51. pp. 555-560. 


Menzel, R. Possibilities of molluscan cultivation in the Caribbean 
1971 symposium on investigations and resources of the Caribbean sea 
and adjacent regions. FAO. Rome, pp. 183*200. 

Mercer, D.E. and Hamilton, L.S. Algunas vent a j as economicas y naturales 

1984 de los ecosistemas de manglares. La Naturaleza y SUB Recursus 
20(2) :14-19. 

Midget, R.J. Coastal aqua culture site selection. In: Shrimp farming 

1985 manual. Ed. Chamberlain, Q.W., Haby, M.G. and Miget, R.J. 
Corpus Christi, Texas, Texas Agri. Exten. Ser, II -1 - II -10. 

MINAGRI/FAQ Znforme de Cuba: Seminario sobre "mane jo integral de 
1984 ecosistemas manglares 11 . Manzanillo, Cuba. 32 p. 

Moore, G.K. Enhancement of Landsat images with diazo color film. Water 
1979 resources development bull. pp. 3-10. 

Morain, S.A. Use of Radar for vegetation analysis. Remote sensing of the 
1976 electromagnetic spectrum. Vol. 3. no. 3. 

M.S.Swaminathan Research Foundation, First Annual Report 1990-1991. 
1992 Centre for Research on Sustainable Agricultural and Rural 
Development. Madras, India. 

Mukherjee, B.B. and Mukherjee, J. Mangroves of Sundarbans, India. 
1978 Phytomorphology, 28(2): 177-191. 

Mulder, N.J. and Hempenius, S.A. Data compression and data reduction 
1974 techniques for the visual interpretation of multispectral ' 
images. ITC Journal, pp. 414-423. 

Murphy, D.H., Meepol, W., and Rau, M.T. Four Papaers on Insects and Ground 
1990 Mesofauna at Ranong. UNDP/UNESCO Regional Mangroves Project 
RAS/ 86/120 Mangrove Occasional Papers no. 7 

Naijung, D. and Mouxun, C. Monitoring Forest Coverage Using Multi- temporal 

1984 Remote Sensing Images pp. 207-211. ESCAP/NRMC Proceedings of 
Regional Conference on multi-level Remote Sensing for Forestry 
Application. 8-12 October 1984. Quezon City, Metro Manila, 

Nayak, S.R., Gupta, M.C. and Chauhan, H.B. Monitoring of Wetland and 

1985 Shoreline on the part of Gujarat Coast Using Landsat pp. 
348-353. Proceedings of the 6th. Asian Conference of Rem.Sens. 
(ACRS) , Asian Association of Remote Sensing, Hyderabad, 1985. 

Nayak, S., Gupta, M.C., Chauhan, H.B., Shaikh, M.G. Pandeya, A. & Rao, R.S. 

1986 Application of Landsat Data for Coastal Zone Monitoring a Case 
Study on the port of West Coast of India. I6p. Paper 
presented at the Regional Seminar on the Application of Remote 
Sensing Techniques to Coastal Zone Management and Environmental 
Monitoring. Dhaka-Bangladesh 18-26 November 1986. 

Ng, F.S.P. Productivity of mangroves systems and replacement systems. 

1987 Published in: Mangrove ecosystems of Asia and the Pacific. Eds 
C.D. Field & A.J. Dartnall. Australian Institute of Marine 
Science , Townsville, Queensland. 

Noakes, D.S.P Notes on the silviculture of the mangrove forests of 

1951 Ma tang, Perak. Malay Forester. 14: 183-196. 

Noakes, D.S.P A working plan for the Matang mangrove forest reserve, 

1952 Perak. Kuala Lumpur, Forest Dept. 173 pp. 

O.D.A. A Forest inventory of the Sundarbans, Bangladesh. Project 

1985 report 140. Land Resources Development Centre, Surrey, England. 


Odum, W.E. Pathways of energy flow in a south Florida estuary. Sea Grant 

1971 Tech. Bull. No, 7. ttiiv. of Miami Sea Grant Program {Living 
Resources) Miami, Florida. 162 pp. 

Odum, W.E. and Heald, E.J. Trophic analyses of an estuarine Mangrove 

1972 community. Bull, of Marine Science, 22. pp. 671-737. 

Odum, W.E. and Heald, E.J. The detritus -based food web of an estuarine 
1975 mangrove community. In: 'Estuarine research', Cronin, L.E., 
Ed. academic Press, Inc. New York. pp. 265-286. 

Odum, W.E., Mclvor, C.C., and Smith, T.J. III. The ecology of the 

1982 mangroves of south Florida: a comunity profile, U.S. Fish and 

Wildlife Service, Office of Biological Services, Washington, 

D.C. FWS/OBS-81/24. 144 pp. 

Ohn, u. Report on Mangrove Forest Products and Utilization of the 
1992 Ayeyarwady Delta. Feasibility Study on Mangrove Reforestation. 
FAQ/FO-.MYA/ 90/003, Field Document No. 4, Yangon. 

Ong, J.E. Coastal zone productivity. In: Kato Kumamoto. I., Mathews, W.H. 
1985 and Ariffin Suhaimi, (eds) . Environl. protection and coastal 
zone Qtgmt. in Asia Pacific, Univ. of Tokyo Press. 244 pp. 

Ong, J.E., Gong, W.K* and Wong, C.H. Productivity and nutrient status of 
1982 litter in a managed mangrove forest. In: Kostermans, A*Y and 
Sastrotomo, Eds. Proceed, of the symp. on mangrove forest 
ecosystem productivity in Southest Asia. Biotrop Special Pub. 
No. 17. 33-41. 

Ong, J.E., Gong, W.K., Wong, C.H. and Dhanarajan, G. Productivity of a 
1979 managed mangrove forest in West Malaysia. In: Proceed, of the 

intl. conf . on trends in applied biology in South East Asia. 

Univ. Sains Malay, pp. 274-284. 

Ong, T.L. Some aspects of trophic relationships of shallow water fishes 
1977 (Selangor Coast). B.Sc. Hons thesis. Univ. Malaya. 55 pp. 

OPSA Manual descriptive del mapa de associaciones de sub-grupos de 
1979 suelos de Costa Rica. 236 p. 

Pannier, F. and Pannier, R.F. Interpretation fisioecologica de la 
1977 distribucion de manglares en las costas del continente 
Suramericano. Intercencia. Vol. 2. No. 3. pp. 153-162 

Pannier, P.F. El efecto de distintas concentraciones sostenas sobre el 
1959 Rhizophora mangle. Acta Cient. Venz. 10: 68-76. 

PCARRD State of the Art Mangrove Research. Forestry Research Series 

1987 No. 4. PCARRD, Los Banos, Laguna, Philippines. 

PCARRD Philippines Recommends for Mangrove Production and Harvesting. 
1992 PCARRD Philippines Recommends Series No. 74, Los Banos, Laguna. 

Pfirez, E.C.M. Inventario forestal en un ecosistema de manglar de la 
1991 Republica de Cuba. Voluntary paper presentaed at the 10th 
World Forestry Congress, 17-26 September, Paris, France. 

Perez P. A. La Apicultura en las zonas de manglar y en el area del Proyecto 

1988 TCP/CUB/8851. Inforroe de Cawpo. Mane jo integrado del 
ecosistema de manglar ", FAO, La Habana, 6 p. 

Pervical, M. and Womersley, J.S. Floristics and ecology of the mangrove 
1975 vegetation of Papua New Guinea. Botany Bull. No 8., Papua Hew 
Guinea, Dept. of Forestry. 


Pool, D.J.i Lugo, A.E.,and Snedaker, S.C. Litter production in mangrove 
1975 forests of southern Florida and Puerto Rico. In: Walsh, Q.S., 

Snedaker, 8. C., and Teas, H., Eds., Proceed of the intl. symp. 

on biol. and mgmt. of mangroves. IFAS, Gainesville, Florida, 

Univ. of Florida, pp. 213-237. 

Pool, D,J., Snedaker, S.C. and Lugo, A.E. Structure of mangrove forests 
1977 in Florida, Puerto Rico, Mexico and Central America. Biotropica 
9: 195-212. 

Potter, T. The problem of fish culture associated with acid sulphate soils 

1977 and methods for their improvement. Contribution to the Joint 
FAO/UNDP/SCPC and SEAFDBC Regional workshop on aqua culture 
engineering. Tigbauan, lloilo, Philippines, 
SCSP/SFDC/77/4En/BP22. 13 pp. 

Pushparajah, M. Application of Remote Sensing to Forestry in Sri-Lanka pp. 

1964 177-179. ESCAP/NRMC Proceedings of Regional Conference on 
multi-level Remote Sensing for Forestry Application. 8-12 
October 1984. Quezon City, Metro Manila, Philippines. 

Putz, F.E. and Chan, H.T Tree growth, dynamics, and productivity in a 
1986 mature mangrove forest in Malaysia. FOR. Ecol. Manage., 

Primack, R.B. and Thomlison, P.B. Sugar secretions from the buds of 

1978 RhizoDhora. Biotropica 10: 74-75. 2 

Qureshi, M.T. Experimental Plantation for Rehabilitation of Mangrove 
1990 Forests in Pakistan. UNDP /UNESCO Regional mangroves project 
RAS/86/120. Mangrove Occasional Papers no. 4. 

Rabinowitz, D. Early growth of mangrove seedlings in Panama and an 
1978 hypotheses concerning the relationship of dispersal and 
zonation. Journal of Biogeography, 5: 113-133. 

Rahman, A., Rahman, N., Hossain, D. and Tugril Hossain, T.I.M. Application 
1986 of Remote Sensing Data in the Studies and .Mapping of Coastal 

Land Accretion and Mangrove Plantations in the Coastal Belt of 
Bangladesh. 14p. Paper presented at the Regional Seminar on the 
Application of Remote Sensing Techniques to Coastal Zone 
Management and Environmental Monitoring. Dhaka-Bangladesh 18-26 
November 1986. 

Ramdial, B.S. Mangrove forests desirable or dispensable. Forestry 
1984 Division Ministry of Agriculture. 24 p. 

Ratanasermpong, S. A study of the Mangrove forest ecosystem and its 
1986 environment by remote sensing, in Proceedings of the Regional 
Seminar on the Application of Remote Sensing techniques to 
.coastal zones management and environment monitoring. 
UNDP/ESCAP and SPARRSO. Dhaka, Bangladesh. 18-26 Nov. 1986. 

Reichert, P. Recent Developments and Trends of Relevance to Multi -Level 
1984 Remote Sensing pp. 219-231. ESCAP/NRMC Proceedings of Regional 
Conference on multi-level Remote Sensing for Forestry 
Application. 8-12 October 1984. Quezon City, Metro Manila, 

Reichle, D.E. Oak Ridge National Laboratory site proposal for analysis of 
1971 the structure and functions of ecosystems in the Deciduous 

Forest Biome. Oak Ridge National Laboratory. Research 

proposal. FY 1972, Oak Ridge, Tenn. 162 pp. 

Remeijn, J.M. Photointerpretation in Forestry. International Institute 
1981 for Aerial Survey and Earth Sciences. ITC, Bnschede, the 
Netherlands . 


Riggs, S.R. Beach geology, JILL J- Clark, The Sanabel report, 
1976 formulation of a comprehensive plan based on natural systems. 
The Conservation Foundation, Washington, D.C. pp. 229-255 

Roberts, E.H. The application of multispectral scanning systems to 

1975 agriculture and forestry. In : Multispectral scanning systems 

and their potential application to earth resources surveys. 

Vol. 4: Earth Science Applications ISRD-CIRS Report. Ed. RAG 

Savigean et Al. pp. 51-83. 

Rollet, B. Introduction 1' Etude des Mangroves du Mexique. Photo - 
1974 interpretation. Types de Ffiret. Bibliographic. Revue Bois et 
Fordts des Tropiques, no. 156. pp. 3-74. 

Rollet, B. Ordenacion integrada de los manglares. Sintesis de siete 
1986 seminaries nationales en America Latina. F.A.O., Rome. 

Rombe, Y.L. Multistage Remote Sensing for Forest Inventory. Application 
1984 with reference to Indonesia, pp. 161-165. ESCAP/NRMC 

Proceedings of Regional Conference on mult i- level Remote 
Sensing for Forestry Application. 8-12 October 1984. Quezon 
City, Metro Manila, Philippines. 

Rosenfield, G.H., Fitzpatrick, K. and Ling, H.S. Sampling for thematic map 

1982 accuracy. Photo. Eng. Rem. Sen. 48 (1) :131-137. 

Saenger, P.,Hegerl, E.J., and Davie, J.D.S. Global Status of mangrove 

1983 ecosystems, Commission on ecology papers No. 3, IUCN, Gland, 
Switzerland, 88 pp. 

SAF Forestry Handbook, 2^ ed. 

Samek, V. Elementos de Silvicultura de los bosques latif olios. Ciencia y 
1974 Tecnica, Institute Cubano del Libro, La Habana. 

San, H.T. and Hong, P.H. The results of investigation of mangrove flora in 

1984 Vietnam. In: The first Vietnam national symposium on mangrove 
ecosystems. Hanoi, pp. 77-86. 

Sandrasegaran, K. Height -Diameter- Age Multiple Regression Models for 
1971 Rhizophora Apiculata Bl. (Syn. Rhizophora Conjugate Linn.) In 

Matang Mangroves, Taiping, West Malaysia. The Malayan 

Forester. Vol. 34 no. 4 pp. 260-275. 

Sasekumar, A. Secondary productivity in mangrove forests. In: 
1984 Productivity of the mangrove ecosystem: Management 

implications. Eds. Ong, J.E. and Gong, W.K. UNESCO/UNDP 

(RAS/79/002/G/01/13) . Penang, pp. 20-28. 

Sasekumar i A. and Lai, J.J. Litter production in three mangrove forest 
1983 zones in the Malay Peninsular. Aquatic Botany, 17: 283-290. 

Schmid-Haas, P. Monitoring change with combined sampling on aerial 
1980 photographs and on the ground. Swiss Federal Institute of 
Forestry Research, pp. 383-388. 

Schumacher, F.X. and Chapman, R.A. Sampling methods in Forestry and Range 
1954 Management. Duke Univ. School of Forestry. Bui. 7, Rev. 
Durham, NC. 221 p. 

SECA/CML Mangroves of Africa and Madagascar. 
1987 Reclamation. SECA/CML/CEC . 

Conservation and 

Seidens ticker, J. and Md. Abdul Hai. The Sundarbans Wildlfe Management 
1983 Plan: Conservation in the Bangladesh coastal zone. WWF/IUCN, 
Gland, Switzerland, 120 pp. 


SEPSA Manual descriptive del mapa geomorfologico de Costa Rica, 
1980 Secretaria Ejecutiva de Planification Sectorial Agropecuaria Y 

de Recursos Naturale* Renovables, Imprenta Nacional, San Jose, 

Costa Rica. 

Set ten, 6. O.K. A note on the growth and yield of Berus. The Malayan 
1953 Forester. Vol. 16 no. 2. pp. 74-87. 

Shahid, M.A. and Pramanik, M.A.H. Application of Remote Sensing Techniques 

1986 to study the relation between the shrimp/fish Farms and 
Mangrove Ecosystem of Bangladesh Coastal Region. 16p. Paper 
presented at the Regional Seminar on the Application of Remote 
Sensing Techniques to Coastal Zone Management and Environmental 
Monitoring. Dhaka -Bangladesh 18-26 November 1986. 

Sicco-Smit, O.S. Sistema de Fotointerpretacion recomendado para los 
1969 bosques humedos tropicales de Colombia. Publicacion del 

Centro Interamericano de Fotointerpretacion, Serie B no. 1. pp. 


Sicco-Smit, O.S. Will the road to the green hell be paved with SLAR? ITC 
1975 Journal, 1975-2:245-266. 

Siddiqi, N.A. and Khan, M.A.S. Two Papers on Mangrove Plantations in 
1990 Bangladesh. UNDP /UNESCO Regional mangroves projects 
RAS/86/120. Mangrove Occasional Papers No. 8. 

Siddiqi, N.A., Islam, M.R., Khan, M.A.S. and Shaidullah, M. Mangrove 
1993 Nurseries in Bangladesh. ISME Mangrove Occasional Papers no.l. 
ISME, Japan. 

Silapathong, C. Remote Sensing application to the study of mangrove 

1983 ecosystem of Thailand. In: ESCAP/UNESCO/NRCT Report on the 
Regional Remote Sensing Training Course of Mangrove Ecosystem 
Bangkok, Nov.28-Dec.16 1983. p. 68-71. 

Silvius, M.J., Chan, H.T. and Shamsudin, I. Evaluation of wetlands of the 

1987 West coast of peninsular Malaysia and their importance for 
natural resource conservation. WWF/FRI /University of 
Malaya /INTERWADER Joint project. WWF Malaysia, Kuala Lumpur. 

Skaley, J.E., Fisher J.R. and Hardy, E.E. A color prediction model for 
1977 imagery analysis. Photo. Eng. Rem. Sen. 43:45-52. 

Snedaker, S.C. and Snedaker, J.G. The mangrove ecosystem: Research 

1984 methods. UNESCO, Paris. 

Soto, R. and Jimenez, J.A. Analisa fisionomico estructural del manglar de 
1982 Puerto Soley, La Cruz, Guanacaste, Costa Rica. Revista de 
Biologia Tropical. 30(2) 151-158. 

Spurr, S.H. Forest Inventory. Ronald Press Company. New York. 476 p. 

Spurr, H.S.; Barnes, B.V. Forest Ecology, John Wiley & SOns, New York, 
1980 687 pp. 

Sribhibadh, A. Status and problems of coastal aquaculture in Thailand. 
1973 In: Pillay, T.V.R. Ed. Coastal aquaculture in the Indo-Pacific 
Region, Fishing News (Books), London, pp. 74-83. 

Srivastava, P.B.L., Ahmad, A.M., Dhanarajan, G. and Hamzah, I. Proceedings 
1979 Symposium on Mangrove & Estuarine Vegetation in Southeast Asia, 
April 25-28, 1978, Selangor, Malaysia. Biotrop Special 
Publication No. 10. Biotrop, Bogor, Indonesia. 


Stern, W.I. and Voight. O.K. Effects of salt concentration on growth of red 
1959 mangrove in culture. Bot. Gaz. 121: 36*39. 

Sugden, E.A.N. and von Cube, H.O. Industrial uses of mangrove (Rhizophora 
1978 racemosa) . In: Poceed. of eighth world forestry congress, 
Jakarta. FID/0-9, 9 pp. 

Sukardjo, S. Some Aspects of Mangrove Ecology. Training Materials for 
1992 Forestry Officers. Integrated Resource Development of the 
Sundarbans Reserved Forest, FAO/UNDP/BGD/84/056, FAO, Rome. 

Sukhatme, P.v. and Sukhatme, B.V. Sampling theory of surveys with 
1970 applications. Iowa State Itaiv, Press. 2nd edition. 491 p. 

Tan, K.F. Some aspects of the biology of the Acetes eiythraceus in Sungai 

1977 Besar, Malaysia. B.Sc. Hons. thesis. Univ. Malaya. 37 pp. 

Tang, H.T. Regeneration stocking adequacy standards. Malay. Forester., Vol. 

1978 41 No. 2. pp. 176-182. 

Taninos Y Maderas Venezolanas C.A. Plan de mane jo forestal de la Unidad 

1974 Norte de los manglares de la Rerserva Forestal de Ouarapiche. 
187 pp. 

Thollot, P. and Kulbicki, M. Overlap between the fish fauna inventories of 
1988 coral reefs, soft bottoms and mangroves in Saint-Vicent Bay 

(New Caledonia) In: Proceedings of the 6th International Coral 
Reef Symposium, Australia, 1988, Vol. 2, pp. 613-616. 

Thorn, B.G. Mangrove ecology and deltaic geomorphology: Tabasco, Mexico. 
1967 Jour. Of Ecol. 55: 301-343. 

Thorn, B.G. Coastal landforms and geomorphic process. In: The mangrove 
1984 ecosystem: research methods. Snedaker, S.C. and Snedaker, J.G. 

Thong, K.L. and A. Sasekumar. The trophic relationship of the fish 

1984 community of the Angsa Bank, Selangor, Malaysia. In: Soepadmo, 
E., Rao, A.N., and Macintosh, D.J. Eds. Proceed, of the 
UNESCO/Asian symp. on mangrove environ: Research and 
management. Univ. Malaya. 385-399. 

Thongchai, C. Study on changes of Mangrove Forest Areas in Thailand by 
1983 using Landdat Imagery* pp. 117-122. 

Tikumponvarkas, S., Fox III, L. and Chaudhury, M.U. Automated mapping of 

1985 forest land use in Tropics. Prepilot study in Southern 
Thailand. In Final Report of UNDP/ESCAP Remote Sensing Program 
Pilot Projects on Forestry in Thailand, The Philippines and 
Indonesia, pp. 13 7 -142. 

Tomlinson, P.B. The botany of mangroves. Cambridge Tropical Biology Series 

1986 Cambridge. Univ. Press. 413 pp. 

Tomlinson, P.B., Primark, R.B. and Bunt, J.S. Preliminary observations on 

1979 floral biology in mangrove Rhizophortceae. Biotropica 11 (4) : 

Turner, R.E. The Segara Anakan Reclamation Project. The Impact on 

1975 Commercial Fisheries. A report to Engineering Consultant Inc., 
Colorado. 71pp. 


Uknali, R.M. and Argete, B.C. A pilot study on the application of raulti- 
1984 level Remote Sensing to forest Resources Mapping in the 

Philippines pp. 111-155. ESCAP/NRMC Proceedings of Regional 
Conference on mult i- level Remote Sensing for Forestry 
Application. 8-12 October 1984. Quezon City, Metro Manila, 

UNESCO The mangrove ecosystem: Research methods. Monographs on 
1984 Oceanographic Methodology. Snedaker and Snedaker, Eds. 

Vantomme, P. Forestry within integrated coastal area management (ICAM) in 
1993 Bast Africa. Paper presented at the Workshop/Conference on 

Integrated Coastal Zone Management in East Africa, Arusha, 21- 

23 April, 1993. FAO, Rome. 

Versoza Jr., C.O. A review of the applications of Remote Sensing to the 

1986 Coastal Zone Resources of the Philippines. 9p. Paper presented 
at the Regional Seminar on the Application of Remote Sensing 
Techniques to Coastal Zone Management and Environmental 
Monitoring. Dhaka -Bangladesh 18-26 November 1986. 

Villalobos, C.S. Animales y plantas comunes de las costas de Costa Rica. 

1982 EUED, Serie Educ. Ambient. No. 6, San Jose. 147 pp. 

Vlcek, J. Videography: Some Remote Sensing Application. Proc. American 

1983 Society of Phot ogramme try. pp 63-69. 

Wacharakitti, S. Mangrove Ecosystem in General pp. 22-33.ESCAP/UNESCO/NRCT 
1983 Regional Remote Sensing Training Course of Mangrove Ecosystem 
Bangkok, Nov.28-Dec.16 1983. 

Walsh, G.E. Mangroves: a review. In: Peinold, R.J. and Queen, W.H., eds. 
1974 Ecology of halophytes. London, Academic Press, Inc., pp. 51-174. 

Walsh, G.E. Exploitation of Mangal. 1. Wet coastal ecosystems. Ed. V.J. 
1977 Chapman, pp. 347-362. 

Walter, H. and Leith, H., Klimadiagramm Weltatlas. Jena. Veb. Gustav 
1960 Fisher Verlag. 

Walter, H and Steiner, M. Die Okologie der ostaafricanischen Mangroven. 
1936 Z. Bot. 30: 65-193. 

Watson, J.D. Mangrove forests of the Malay Peninsula. Fraser & Neave, 
1928 Malayan Forest Records, Vol. 6. Singapore, 275 pp. 

Watts, J.C.D. Phosphate retention in acid-sulphate pond muds from Malacca 
1969 area. Malay. Agric. Jour. 47: 187-202. 

Weechakit, D. Growth and survival of private mangrove plantations 

1987 (Rhizophora apiculata) at Amphoe Amphawa, Samut Province. M.Sc. 
Thesis. Kasetsart University Graduate School, Thailand. 

Weir, M.J.C. An assessment of simple plotting instruments for resource 
1981 mapping. Paper presented at a Conference on Matching Remote 
Sensing technologies and their application convened in London 
by the Remote Sensing Society. 16-18 Dec. 1981. pp. 223-249. 

Weir, M.J.C. Mapping for resources surveys. ITC. Bnschede, The 
1983 Netherlands. 63p. 

White, K.J. (ed.) Research Considerations in Coastal Afforestation. Field 
1979 Document No. 16. UNDP/FAO Project BGD/72/005 and the Forest 
Research Institute, Chittagong, Bangladesh. 120pp. 


Woodcock, W.E. Aerial Reconnaissance and Photogrammetry with small 
1976 cameras* Photogrammetric Engineering and Remote Sensing. Vol. 
42, no. 4. pp. 503-511. 

Wong, K.W. Geometric and cartographic accuracy of ERTS-1 imagery. 
1975 Photo. Eng. Rem. Sen. 41:621-635. 

Wyrtki, K. Physical oceanography of the Southeast Asian waters. Naga 
1961 Report, Vol. 2. Scientific Results of Marine Investigations in 

the South China Sea and the Gulf of Thailand, 1959-1961. Univ. 

of California, La Jolla, Cal., 195 pp. 

Yeo, A.R. and Flower, T.J. Salt tolerance in the halophyte Suaeda 
1980 maritima L. Dum: Evaluation of the effect of salinity upon 
growth. Jour. Exp. Bot. 31: 117-1183. 

Young, T.R. Legal and institutional issues in the creation and 
1993 implementation of integrated coastal zone management programs 
in East Africa. Paper presented at the Workshop/Conference on 
Integrated Coastal Zone Management in East Africa, Arusha, 
Tanzania 21-23 April, 1993. FAO, Rome. 





3 Rv,2 





1 1 







17 Sup.1 

17 Sup.2 








Foreat utilization oontreots on fwblic^ 

Planning forttt roadt and Iwvttting systems, 1977 


Worid litt of fortttrytchooli, 1977 (E/F/S) 

World litt of fortftrytchooli, 1981 (E/F/S) 

World Hat of fortitry schools, 1986 (E/F/S) 

World pulp and papar damand, lupply and trada 

-Vol. 1,1977 (EFS) 

World pulp and papar demand, aupply and trada 

-Vol. 2, 1977 (EFS) 

Tha markatlng of tropical wood, 1976 (E S) 

National parki planning, 1976 (E F S**) 

Fortttry for local community development, 1978 


Eatabllihmant technique! for forest plantations, 1978 


Wood chips - production, handling, transport, 1976 


Assessment of logging costs from forest inventories In 

the tropics 

- 1, Principles and methodology, 1978 (EFS) 
Assessment of logging costs from forest Inventories in 
the tropics 

- 2. Data collection and calculations, 1978 (E F S) 
Savanna afforestation In Africa, 1977 (E F) 
China: forestry support for agriculture, 1978 (E) 
Forest products prices 1960-1977, 1979 (E/F/S) 
Mountain forest roads and harvesting, 1979 (E) 
Logging and transport In steep terrain, 1985 (E) 
AGRIS forestry - world catalogue of information and 
documentation services, 1979 (E/F/S) 

China: Integrated wood processing industries, 1979 

Economic analysis of forestry projects, 1979 (EFS) 

Economic analysis of forestry projects: case studies, 

1979 (E S) 

Economic analysis of forestry projects: readings, 1980 


Forest products prices 1960-1978, 1980 (E/F/S) 

Pulping and paper-making properties of fast-growing 

plantation wood species 

- Vol. 1, 1980 (E) 

Pulping and paper-making properties of fast-growing 
plantation wood species 

- Vol. 2, 1980 (E) 

Forest tree Improvement, 1985 (C EFS) 

A guida to forest aaad handling, 1985 (E S ) 

Impact on soils of fast^rowingspedes in lowland humid 

tropics, 1980 (EFS) 

Forest volume estimation and yield prediction 

- Vol. 1. Volume estimation, 1980 (C E F S) 
Forest volume estimation and yield prediction 

- Vol. 2. Yield prediction, 1980 (C E F S) 
Forest products prices 1961-1980, 1981 (E/F/S) 
Cable logging systems, 1981 (C E) 
Publlcfofestry administrations In Latin America, 1981 (E) 
Forestry and rural development, 1981 (EFS) 
Manual of forest Inventory, 1981 (E F) 

Small and medium sawmills In developing countries, 
1981 (ES) 

29 World fomt product!, dwnmdanliupply1<N0200a 
1962 (EFS) 

30 TnpioalfortttrnouroM, 1962 (EFS) 

31 Appropriate technology to forwtry, 1982 (E) 

32 QaMHIetfon and dtfinHJoni of form product*, 1962 

33 Logging of mountain fornte, 1982 (EFS) 

34 Fruit-owing fond (ran, 1962 (EPS) 

35 FbTMlry In China, 19B2(CE) 

36 Batlc technology In fomt opwateni, 1962 (EFS) 

37 Conservation and development of tropical forest 
resources, 1982 (EFS) 

38 Forest products prices 1962*1981, 1982 (E/F/S) 

39 Frame aaw manual, 1982 (E) 

40 Circular aaw manual, 1983 (E) 

41 Simple technologies for charcoal making, 1983 (E F S) 

42 Fuetwood supplies in the developing bounties, 1983 

43 Forest revenue systems In developing countries, 1983 

44/1 Food and fruit-bearing forest species 

- 1, Examples from eastern Africa, 1983 (E F S) 
44/2 Food and fruit-bearing forest spedes 

- 2. Examples from southeastern Asia, 1984 (EFS) 
44/3 Food and fruit-bearing forest species 

- 3. Examples from Latin America, 1986 (E S) 

45 Establishing pulp and paper mills, 1983 (E) 

46 Forest products prices 1963-1982, 1983 (E/F/S) 

47 Technical forestry education - design and 
implementation, 1984 (EFS) 

48 Land evaluation for forestry, 1984 (C EFS) 

49 Wood extraction with oxen and agricultural tractors, 
1986 (EFS) 

50 Changes In shifting cultivation In Africa, 1984 (E F) 
50/1 Changes In shifting cultivation In Africa - seven 

case-studies, 1985 (E) 
51/1 Studies on the volume and yield of tropical forest 

stands - 1. Dry forest formations, 1989 (E F) 
52/1 Cost estimating In eawmilling industries: guidelines, 

1984 (E) 
52/2 Reid manual on coat estimation In lawmllllng 

industries, 1985 (E) 

53 Intensive multiple-use forest management In Kerala, 

54 PlanHlcacl6ndeldevwrolloforeetalJ964(S) 

55 Intensive multiple-use forest management In the 
tropics, 1985 (EFS) 

56 Breeding poplars to disease resistance, 1985 (E) 

57 Coconut wood - Processing and use, 1985 (ES) 

58 Sawdoctoring manual, 1985 (E S) 

59 The ecological effects of eucalyptus 1965 (C EFS) 

60 Monitoring and evaluation of participatory forestry 
projects, 1985 (EFS) 

61 Forest products prices 1965-1984, 1965 (E/F/S) 

62 World list of institutions engaged In totetryandtoreet 
products research, 1985 (E/F/S) 

63 Industrial charcoal making, 1965 

64 Tree growing by rural people, 1965 (ArEFS) 

65 Fbrest legislation In selected African countries, 1966 

3 So 























Foratiry **< ixganixaflon, 1006 (C E 8) 
8onw mtdfeiMl fomft plants of Mrioa and Ltttn 

Appropriate fentt InduatriM, 1086 (E) 

NMMQMMflt Of IQfMl MdUMIMi 1000 |E| 

reeearoh Institutions, 1906 (E/F/S) 

Wood gas as engine fuel, 1966 (ES) 

Forest products: world outiook projections 1965-2000, 

1066 (E/F/S) 

i processing, 1966 (E) 

of tooial forwtry in Mli, 1966 (E) 
Wood pretwvitfofi nfMMHMH, 1066 (E) 
Ditttbook on ofKwnQMWi IFM Md shrub spools And 
, 1966 (E) 

Smill^Ottto fonit butd pfociiiiinn initrpfiiti, 1967 





1902 (E) 

management - Principles and concepts, 1903 (E) 
A decade of wood energy activities wttWn the Nabob! 
Programme cf Action, 1903 (E) 
Directory of forestry research organizations, 1903 (E) 
Proceedings of the Meeting of Experts on Forestry 
Research, 1993 (E/F/S) 

bhflMMhtfl -* ** *-* t* A|*A ^* - B*^ r^f>lM AMAk^bl* 

rorvevy ponoiee m me near BBM region ^naiym 

and syntheeis, 1903 (E) 

Forest reeouroee assessment 1990 -Tropical 

countries, 1903 (E) 

Br $hu storage of seeds, poNon and <n vMro cultures of 

perennial woody plant species, 1903 (E) 

Assessing forestry project Impacts: Issues and 

Forwfry wctentton molhodi, 1967 (E) 
QuidsHntt for fonit policy formulttkxi, 1987 (CE) 
Fort* produeto prloit 1967-1966, 1968 (E/F/S) 
Tfftdo In focttt pfoduott! 6 study of ths bsnisrs ftosd 
by tht dovolopino ooMntrtts, 1966 (E) 
Forsst produoti: Wbrtd outlook projections 1 967-2000 
- Product and country tib*, 1968 (E/F/S) 
Forsstry txlsnsion ourrioult, 1686 (E/F/8) 
Forsstry pofcits In Europe, 1968 (E) 
SfTu^sottls hsrvs sting optritkxis of wood shd 
norvwood forost products Involving rural poopto, 1968 

of tropical inoisi forasis in Afnoa, 1969 


Forestry policies of selected countries In Asia and the 
Pacific, 1903 (E) 

Les panneaux A base de bole, 1903 (F) 
Mangrove forest management guidelines, 1993 (E) 

Availability: December 1993 



- Arabic 

- Chinese 



Out of print 
In preparation 

- Spanish 

Tfce fi 40 Tectafca/ Papen are auaM* through fa tuthortztd 

nt systsms of tropical Asia, 


Rsviaw of f 

1969 (E) 

Fbrtstry and food saourlty, 1969 (Ar E 8) 

Ossign manual on baste wood harvesting technology, 

1969 (EPS) 

(Pubttshad only as FAQ Training Series, No. 18) 

Forestry policies In Europe -An analysis, 1969 (E) 

Energy conservation In the mechanical forest 

Industries, 1990 (E S) 

Manual on sawmi operational maintenance, 

1900 (E) 

Forest products prices 1969-1968, 1990 (E/F/S) 

Planning and managing foreetry research: guidelines 

for manners, 1990 (E) 

Nonfood forest products: the way ahead, 1991 (E 8) 


FAO, VKedefle Terme df Camca/te, 00100 Rom*, My. 

Colt control In forest harvesting and road construction, 

1962 (E) 

Introduction to ergonomics In forestry In developing 

oourrtries, 1992(E) 

Management and conservation of dosed forests in 

i In fomky, 1001 (E) 

t pr 

ttttt fl*l 

ucts price 

1971-1990, 1902 (E) 

Qompetidium cf pulp 6nd paper training and reeearoh